root/other-projects/nightly-tasks/diffcol/trunk/model-collect/Word-PDF-Formatting/archives/HASHeaa2.dir/doc.xml @ 29405

Revision 29405, 245.9 KB (checked in by ak19, 5 years ago)

Trying to rebuild the Word-PDF-Formatting collection with unique dc.Title metadata for docs that have identical final names and with the 2nd browsing classifier sorted on dc.Title, in order to produce a consistent order for browse classifiers' children (a consistent presentation order of the files under the browsing classifiers). This is necessary for perl 5.18/5.17 and later, since they randomise the order of children of unsorted classifiers and for those children with identical filenames. Changes made particularly to collect.cfg and import/metadata.xml

Line 
1<?xml version="1.0" encoding="utf-8" standalone="no"?>
2<!DOCTYPE Archive SYSTEM "http://greenstone.org/dtd/Archive/1.0/Archive.dtd">
3<Archive>
4<Section>
5  <Description>
6    <Metadata name="gsdldoctype">indexed_doc</Metadata>
7    <Metadata name="Language">en</Metadata>
8    <Metadata name="Encoding">utf8</Metadata>
9    <Metadata name="GENERATOR">wvWare/wvWare version 1.2.4</Metadata>
10    <Metadata name="Title">1997-00 Listing of Working Papers</Metadata>
11    <Metadata name="URL">http://Scratch/ak19/gs2-svn-22Oct2014/collect/Word-PDF-Formatting/tmp/1414470427_1/word01.html</Metadata>
12    <Metadata name="UTF8URL">http://Scratch/ak19/gs2-svn-22Oct2014/collect/Word-PDF-Formatting/tmp/1414470427_1/word01.html</Metadata>
13    <Metadata name="gsdlsourcefilename">import/word01.doc</Metadata>
14    <Metadata name="gsdlconvertedfilename">tmp/1414470427_1/word01.html</Metadata>
15    <Metadata name="OrigSource">word01.html</Metadata>
16    <Metadata name="Source">word01.doc</Metadata>
17    <Metadata name="SourceFile">word01.doc</Metadata>
18    <Metadata name="Plugin">WordPlugin</Metadata>
19    <Metadata name="FileSize">110080</Metadata>
20    <Metadata name="FilenameRoot">word01</Metadata>
21    <Metadata name="FileFormat">Word</Metadata>
22    <Metadata name="srcicon">_icondoc_</Metadata>
23    <Metadata name="srclink_file">doc.doc</Metadata>
24    <Metadata name="srclinkFile">doc.doc</Metadata>
25    <Metadata name="Identifier">HASHeaa2992e081949673150f3</Metadata>
26    <Metadata name="lastmodified">1414470425</Metadata>
27    <Metadata name="lastmodifieddate">20141028</Metadata>
28    <Metadata name="oailastmodified">1414470427</Metadata>
29    <Metadata name="oailastmodifieddate">20141028</Metadata>
30    <Metadata name="assocfilepath">HASHeaa2.dir</Metadata>
31    <Metadata name="gsdlassocfile">doc.doc:application/msword:</Metadata>
32  </Description>
33  <Content>
34
35&lt;!--Section Begins--&gt;&lt;br&gt;
36
37
38
39&lt;p&gt;&lt;div name=&quot;Title&quot; align=&quot;center&quot; style=&quot;margin: 0.00mm -0.12mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
40
41&lt;p style=&quot;text-indent: 0.00mm; text-align: center; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
42&lt;b&gt;1997-00 Listing of Working Papers &lt;/b&gt;
43&lt;/p&gt;&lt;/div&gt;
44
45
46
47&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
48
49&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
50
51&lt;/p&gt;&lt;/div&gt;
52
53
54
55&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
56
57&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
58
59&lt;/p&gt;&lt;/div&gt;
60
61
62
63&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
64
65&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
66
67&lt;/p&gt;&lt;/div&gt;
68
69
70
71&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
72
73&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
742000/1
75&lt;/p&gt;&lt;/div&gt;
76
77
78
79&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
80
81&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
82Using compression to identify acronyms in text
83&lt;/p&gt;&lt;/div&gt;
84
85
86
87&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
88
89&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
90Stuart Yeates, David Bainbridge, Ian H. Witten
91&lt;/p&gt;&lt;/div&gt;
92
93
94
95&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
96
97&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
98Text mining is about looking for patterns in natural language text, and may be defined as the process of analyzing text to extract information from it for particular purposes.  In previous work, we claimed that compression is a key technology for text mining, and backed this up with a study that showed how particular kinds of lexical tokens&amp;mdash;names, dates, locations, &lt;i&gt;etc.&lt;/i&gt;&amp;mdash;can be identified and located in running text, using compression models to provide the leverage necessary to distinguish different token types (Witten &lt;i&gt;et al.&lt;/i&gt;, 1999)
99&lt;/p&gt;&lt;/div&gt;
100
101
102
103&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
104
105&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
106
107&lt;/p&gt;&lt;/div&gt;
108
109
110
111&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
112
113&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1142000/2
115&lt;/p&gt;&lt;/div&gt;
116
117
118
119&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
120
121&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
122Text categorization using compression models
123&lt;/p&gt;&lt;/div&gt;
124
125
126
127&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
128
129&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
130Eibe Frank, Chang Chui, Ian H. Witten
131&lt;/p&gt;&lt;/div&gt;
132
133
134
135&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
136
137&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
138Text categorization, or the assignment of natural language texts to predefined categories based on their content, is of growing importance as the volume of information available on the internet continues to overwhelm us.  The use of predefined categories implies a &amp;ldquo;supervised learning&amp;rdquo; approach to categorization, where already-classified articles - which effectively define the categories - are used as &amp;ldquo;training data&amp;rdquo; to build a model that can be used for classifying new articles that comprise the &amp;ldquo;test data&amp;rdquo;.  This contrasts with &amp;ldquo;unsupervised&amp;rdquo; learning, where there is no training data and clusters of like documents are sought amongst the test articles.  With supervised learning, meaningful labels (such as keyphrases) are attached to the training documents, and appropriate labels can be assigned automatically to test documents depending on which category they fall into.
139&lt;/p&gt;&lt;/div&gt;
140
141
142
143&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
144
145&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
146
147&lt;/p&gt;&lt;/div&gt;
148
149
150
151&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
152
153&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1542000/3
155&lt;/p&gt;&lt;/div&gt;
156
157
158
159&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
160
161&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
162Reserved for Sally Jo
163&lt;/p&gt;&lt;/div&gt;
164
165
166
167&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
168
169&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
170
171&lt;/p&gt;&lt;/div&gt;
172
173
174
175&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
176
177&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1782000/4
179&lt;/p&gt;&lt;/div&gt;
180
181
182
183&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
184
185&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
186Interactive machine learning&amp;mdash;letting users build classifiers
187&lt;/p&gt;&lt;/div&gt;
188
189
190
191&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
192
193&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
194Malcolm Ware, Eibe Frank, Geoffrey Holmes, Mark Hall, Ian H. Witten
195&lt;/p&gt;&lt;/div&gt;
196
197
198
199&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
200
201&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
202According to standard procedure, building a classifier is a fully automated process that follows data preparation by a domain expert.  In contrast, &amp;lt;I&amp;gt;interactive&amp;lt;/I&amp;gt;machine learning engages users in actually generating the classifier themselves.  This offers a natural way of integrating background knowledge into the modeling stage&amp;mdash;so long as interactive tools can be designed that support efficient and effective communication.  This paper shows that appropriate techniques can empower users to create models that compete with classifiers built by state-of-the-art learning algorithms.  It demonstrates that users&amp;mdash;even users who are not domain experts&amp;mdash;can often construct good classifiers, without any help from a learning algorithm, using a simple two-dimensional visual interface.  Experiments demonstrate that, not surprisingly, success hinges on the domain: if a few attributes can support good predictions, users generate accurate classifiers, whereas domains with many high-order attribute interactions favor standard machine learning techniques.  The future challenge is to achieve a symbiosis between human user and machine learning algorithm.
203&lt;/p&gt;&lt;/div&gt;
204
205
206
207&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
208
209&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
210
211&lt;/p&gt;&lt;/div&gt;
212
213
214
215&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
216
217&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2182000/5
219&lt;/p&gt;&lt;/div&gt;
220
221
222
223&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
224
225&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
226KEA: Practical automatic keyphrase extraction
227&lt;/p&gt;&lt;/div&gt;
228
229
230
231&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
232
233&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
234Ian H. Witten, Gordon W. Paynter, Eibe Frank, Carl Gutwin, Craig G. Nevill-Manning
235&lt;/p&gt;&lt;/div&gt;
236
237
238
239&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
240
241&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
242Keyphrases provide semantic metadata that summarize and characterize documents.  This paper describes Kea, an algorithm for automatically extracting keyphrases from text.  Kea identifies candidate keyphrases using lexical methods, calculates feature values for each candidate, and uses a machine learning algorithm to predict which candidates are good keyphrases.  The machine learning scheme first builds a prediction model using training documents with known keyphrases, and then uses the model to find keyphrases in new documents.  We use a large test corpus to evaluate Kea's effectiveness in terms of how many author-assigned keyphrases are correctly identified.  The system is simple, robust, and publicly available.
243&lt;/p&gt;&lt;/div&gt;
244
245
246
247&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
248
249&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
250
251&lt;/p&gt;&lt;/div&gt;
252
253
254
255&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
256
257&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2582000/6
259&lt;/p&gt;&lt;/div&gt;
260
261
262
263&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
264
265&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
266&lt;i&gt;ÎŒ&lt;/i&gt;-Charts and Z:  hows, whys and wherefores
267&lt;/p&gt;&lt;/div&gt;
268
269
270
271&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
272
273&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
274Greg Reeve, Steve Reeves
275&lt;/p&gt;&lt;/div&gt;
276
277
278
279&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
280
281&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
282In this paper we show, by a series of examples, how the &lt;i&gt;ÎŒ&lt;/i&gt;-chart formalism can be translated into Z.  We give reasons for why this is an interesting and sensible thing to do and what it might be used for.
283&lt;/p&gt;&lt;/div&gt;
284
285
286
287&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
288
289&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
290
291&lt;/p&gt;&lt;/div&gt;
292
293
294
295&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
296
297&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
298
299&lt;/p&gt;&lt;/div&gt;
300
301
302
303&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
304
305&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3062000/7
307&lt;/p&gt;&lt;/div&gt;
308
309
310
311&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
312
313&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
314One dimensional non-uniform rational B-splines for animation control
315&lt;/p&gt;&lt;/div&gt;
316
317
318
319&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
320
321&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
322Abdelaziz Mahoui
323&lt;/p&gt;&lt;/div&gt;
324
325
326
327&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
328
329&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
330Most 3D animation packages use graphical representations called motion graphs to represent the variation in time of the motion parameters.  Many use two-dimensional B-splines as animation curves because of their power to represent free-form curves.  In this project, we investigate the possibility of using One-dimensional Non-Uniform Rational B-Spline (NURBS) curves for the interactive construction of animation control curves.  One-dimensional NURBS curves present the potential of solving some problems encountered in motion graphs when two-dimensional B-splines are used.  The study focuses on the properties of One-dimensional NURBS mathematical model.  It also investigates the algorithms and shape modification tools devised for two-dimensional curves and their port to the One-dimensional NURBS model.  It also looks at the issues related to the user interface used to interactively modify the shape of the curves.
331&lt;/p&gt;&lt;/div&gt;
332
333
334
335&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
336
337&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
338
339&lt;/p&gt;&lt;/div&gt;
340
341
342
343&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
344
345&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3462000/8
347&lt;/p&gt;&lt;/div&gt;
348
349
350
351&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
352
353&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
354Correlation-based feature selection of discrete and numeric class machine learning
355&lt;/p&gt;&lt;/div&gt;
356
357
358
359&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
360
361&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
362Mark A. Hall
363&lt;/p&gt;&lt;/div&gt;
364
365
366
367&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
368
369&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
370Algorithms for feature selection fall into two broad categories: &amp;lt;I&amp;gt;wrappers&amp;lt;/I&amp;gt;that use the learning algorithm itself to evaluate the usefulness of features and &amp;lt;I&amp;gt;filters&amp;lt;/I&amp;gt;that evaluate features according to heuristics based on general characteristics of the data.  For application to large databases, filters have proven to be more practical than wrappers because they are much faster.  However, most existing filter algorithms only work with discrete classification problems.  This paper describes a fast, correlation-based filter algorithm that can be applied to continuous and discrete problems.  The algorithm often out-performs the well-known ReliefF attribute estimator when used as a preprocessing step for naïve Bayes, instance-based learning, decision trees, locally weighted regression, and model trees.  It performs more feature selection than ReliefF does-reducing the data dimensionality by fifty percent in most cases.  Also, decision and model trees built from the prepocessed data are often significantly smaller.
371&lt;/p&gt;&lt;/div&gt;
372
373
374
375&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
376
377&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
378
379&lt;/p&gt;&lt;/div&gt;
380
381
382
383&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
384
385&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3862000/9
387&lt;/p&gt;&lt;/div&gt;
388
389
390
391&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
392
393&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
394A development environment for predictive modelling in foods
395&lt;/p&gt;&lt;/div&gt;
396
397
398
399&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
400
401&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
402G. Holmes, M.A. Hall
403&lt;/p&gt;&lt;/div&gt;
404
405
406
407&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
408
409&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
410WEKA (Waikato Environment for Knowledge Analysis) is a comprehensive suite of Java class libraries that implement many state-of-the-art machine learning/data mining algorithms.  Non-programmers interact with the software via a user interface component called the Knowledge Explorer.
411&lt;/p&gt;&lt;/div&gt;
412
413
414
415&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
416
417&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
418
419&lt;/p&gt;&lt;/div&gt;
420
421
422
423&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
424
425&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
426Applications constructed from the WEKA class libraries can be run on any computer with a web browsing capability, allowing users to apply machine learning techniques to their own data regardless of computer platform.  This paper describes the user interface component of the WEKA system in reference to previous applications in the predictive modeling of foods.
427&lt;/p&gt;&lt;/div&gt;
428
429
430
431&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
432
433&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
434
435&lt;/p&gt;&lt;/div&gt;
436
437
438
439&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
440
441&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4422000/10
443&lt;/p&gt;&lt;/div&gt;
444
445
446
447&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
448
449&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
450Benchmarking attribute selection techniques for data mining
451&lt;/p&gt;&lt;/div&gt;
452
453
454
455&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
456
457&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
458Mark A. Hall, Geoffrey Holmes
459&lt;/p&gt;&lt;/div&gt;
460
461
462
463&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
464
465&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
466Data engineering is generally considered to be a central issue in the development of data mining applications.  The success of many learning schemes, in their attempts to construct models of data, hinges on the reliable identification of a small set of highly predictive attributes.  The inclusion of irrelevant, redundant and noisy attributes in the model building process phase can result in poor predictive performance and increased computation.
467&lt;/p&gt;&lt;/div&gt;
468
469
470
471&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
472
473&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
474
475&lt;/p&gt;&lt;/div&gt;
476
477
478
479&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
480
481&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
482Attribute selection generally involves a combination of search and attribute utility estimation plus evaluation with respect to specific learning schemes.  This leads to a large number of possible permutations and has led to a situation where very few benchmark studies have been conducted.
483&lt;/p&gt;&lt;/div&gt;
484
485
486
487&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
488
489&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
490
491&lt;/p&gt;&lt;/div&gt;
492
493
494
495&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
496
497&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
498This paper presents a benchmark comparison of several attribute selection methods.  All the methods produce an attribute ranking, a useful devise of isolating the individual merit of an attribute.  Attribute selection is achieved by cross-validating the rankings with respect to a learning scheme to find the best attributes.  Results are reported for a selection of standard data sets and two learning schemes C4.5 and naïve Bayes.
499&lt;/p&gt;&lt;/div&gt;
500
501
502
503&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
504
505&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
506
507&lt;/p&gt;&lt;/div&gt;
508
509
510
511&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
512
513&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
5142000/11
515&lt;/p&gt;&lt;/div&gt;
516
517
518
519&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
520
521&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
522Steve Reeves, Greg Reeve
523&lt;/p&gt;&lt;/div&gt;
524
525
526
527&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
528
529&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
530
531&lt;/p&gt;&lt;/div&gt;
532
533
534
535&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
536
537&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
538
539&lt;/p&gt;&lt;/div&gt;
540
541
542
543&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
544
545&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
5462000/12
547&lt;/p&gt;&lt;/div&gt;
548
549
550
551&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
552
553&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
554Malika Mahoui, Sally Jo Cunningham
555&lt;/p&gt;&lt;/div&gt;
556
557
558
559&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
560
561&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
562Transaction logs are invaluable sources of fine-grained information about users' search behavior.  This paper compares the searching behavior of users across two WWW-accessible digital libraries: the New Zealand Digital Library's Computer Science Technical Reports collection (CSTR), and the Karlsruhe Computer Science Bibliographies (CSBIB) collection.  Since the two collections are designed to support the same type of users-researchers/students in computer science a comparative log analysis is likely to uncover common searching preferences for that user group.  The two collections differ in their content, however; the CSTR indexes a full text collection, while the CSBIB is primarily a bibliographic database.  Differences in searching behavior between the two systems may indicate the effect of differing search facilities and content type.
563&lt;/p&gt;&lt;/div&gt;
564
565
566
567&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
568
569&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
570
571&lt;/p&gt;&lt;/div&gt;
572
573
574
575&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
576
577&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
578
579&lt;/p&gt;&lt;/div&gt;
580
581
582
583&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
584
585&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
586
587&lt;/p&gt;&lt;/div&gt;
588
589
590
591&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
592
593&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
594
595&lt;/p&gt;&lt;/div&gt;
596
597
598
599&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
600
601&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
602
603&lt;/p&gt;&lt;/div&gt;
604
605
606
607&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
608
609&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
610
611&lt;/p&gt;&lt;/div&gt;
612
613
614
615&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
616
617&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
61899/1
619&lt;/p&gt;&lt;/div&gt;
620
621
622
623&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
624
625&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
626Lexical attraction for text compression
627&lt;/p&gt;&lt;/div&gt;
628
629
630
631&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
632
633&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
634Joscha Bach, Ian H. Witten
635&lt;/p&gt;&lt;/div&gt;
636
637
638
639&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
640
641&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
642New methods of acquiring structural information in text documents may support better compression by identifying an appropriate prediction context for each symbol.  The method of &amp;ldquo;lexical attraction&amp;rdquo; infers syntactic dependency structures from statistical analysis of large corpora.  We describe the generation of a lexical attraction model, discuss its application to text compression, and explore its potential to outperform fixed-context models such as word-level PPM.  Perhaps the most exciting aspect of this work is the prospect of using compression as a metric for structure discovery in text.
643&lt;/p&gt;&lt;/div&gt;
644
645
646
647&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
648
649&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
650
651&lt;/p&gt;&lt;/div&gt;
652
653
654
655&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
656
657&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
658
659&lt;/p&gt;&lt;/div&gt;
660
661
662
663&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
664
665&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
66699/2
667&lt;/p&gt;&lt;/div&gt;
668
669
670
671&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
672
673&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
674Generating rule sets from model trees
675&lt;/p&gt;&lt;/div&gt;
676
677
678
679&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
680
681&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
682Geoffrey Holmes, Mark Hall, Eibe Frank
683&lt;/p&gt;&lt;/div&gt;
684
685
686
687&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
688
689&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
690Knowledge discovered in a database must be represented in a form that is easy to understand.  Small, easy to interpret nuggets of knowledge from data are one requirement and the ability to induce them from a variety of data sources is a second.  The literature is abound with classification algorithms, and in recent years with algorithms for time sequence analysis, but relatively little has been published on extracting meaningful information from problems involving continuous classes (regression).
691&lt;/p&gt;&lt;/div&gt;
692
693
694
695&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
696
697&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
698
699&lt;/p&gt;&lt;/div&gt;
700
701
702
703&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
704
705&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
706Model trees-decision trees with linear models at the leaf nodes-have recently emerged as an accurate method for numeric prediction that produces understandable models.  However, it is well known that decision lists-ordered sets of If-Then rules-have the potential to be more compact and therefore more understandable than their tree counterparts.
707&lt;/p&gt;&lt;/div&gt;
708
709
710
711&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
712
713&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
714
715&lt;/p&gt;&lt;/div&gt;
716
717
718
719&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
720
721&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
722In this paper we present an algorithm for inducing simple, yet accurate rule sets from model trees.  The algorithm works by repeatedly building model trees and selecting the best rule at each iteration.  It produces rule sets that are, on the whole, as accurate but smaller than the model tree constructed from the entire dataset.  Experimental results for various heuristics which attempt to find a compromise between rule accuracy and rule coverage are reported.  We also show empirically that our method produces more accurate and smaller rule sets than the commercial state-of-the-art rule learning system Cubist.
723&lt;/p&gt;&lt;/div&gt;
724
725
726
727&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
728
729&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
730
731&lt;/p&gt;&lt;/div&gt;
732
733
734
735&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
736
737&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
738
739&lt;/p&gt;&lt;/div&gt;
740
741
742
743&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
744
745&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
74699/3
747&lt;/p&gt;&lt;/div&gt;
748
749
750
751&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
752
753&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
754A diagnostic tool for tree based supervised classification learning algorithms
755&lt;/p&gt;&lt;/div&gt;
756
757
758
759&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
760
761&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
762Leonard Trigg, Geoffrey Holmes
763&lt;/p&gt;&lt;/div&gt;
764
765
766
767&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
768
769&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
770The process of developing applications of machine learning and data mining that employ supervised classification algorithms includes the important step of knowledge verification.  Interpretable output is presented to a user so that they can verify that the knowledge contained in the output makes sense for the given application.  As the development of an application is an iterative process it is quite likely that a user would wish to compare models constructed at various times or stages.
771&lt;/p&gt;&lt;/div&gt;
772
773
774
775&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
776
777&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
778
779&lt;/p&gt;&lt;/div&gt;
780
781
782
783&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
784
785&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
786One crucial stage where comparison of models is important is when the accuracy of a model is being estimated, typically using some form of cross-validation.  This stage is used to establish an estimate of how well a model will perform on unseen data.  This is vital information to present to a user, but it is also important to show the degree of variation between models obtained from the entire dataset and models obtained during cross-validation.  In this way it can be verified that the cross-validation models are at least structurally aligned with the model garnered from the entire dataset.
787&lt;/p&gt;&lt;/div&gt;
788
789
790
791&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
792
793&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
794
795&lt;/p&gt;&lt;/div&gt;
796
797
798
799&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
800
801&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
802This paper presents a diagnostic tool for the comparison of tree-based supervised classification models.  The method is adapted from work on approximate tree matching and applied to decision trees.  The tool is described together with experimental results on standard datasets.
803&lt;/p&gt;&lt;/div&gt;
804
805
806
807&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
808
809&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
810
811&lt;/p&gt;&lt;/div&gt;
812
813
814
815&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
816
817&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
818
819&lt;/p&gt;&lt;/div&gt;
820
821
822
823&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
824
825&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
82699/4
827&lt;/p&gt;&lt;/div&gt;
828
829
830
831&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
832
833&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
834Feature selection for discrete and numeric class machine learning
835&lt;/p&gt;&lt;/div&gt;
836
837
838
839&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
840
841&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
842Mark A. Hall
843&lt;/p&gt;&lt;/div&gt;
844
845
846
847&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
848
849&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
850Algorithms for feature selection fall into two broad categories: &amp;lt;I&amp;gt;wrappers&amp;lt;/I&amp;gt;use the learning algorithm itself to evaluate the usefulness of features, while &amp;lt;I&amp;gt;filters&amp;lt;/I&amp;gt;evaluate features according to heuristics based on general characteristics of the data.  For application to large databases, filters have proven to be more practical than wrappers because they are much faster.  However, most existing filter algorithms only work with discrete classification problems.
851&lt;/p&gt;&lt;/div&gt;
852
853
854
855&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
856
857&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
858
859&lt;/p&gt;&lt;/div&gt;
860
861
862
863&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
864
865&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
866This paper describes a fast, correlation-based filter algorithm that can be applied to continuous and discrete problems.  Experiments using the new method as a preprocessing step for naïve Bayes, instance-based learning, decision trees, locally weighted regression, and model trees show it to be an effective feature selector- it reduces the data in dimensionality by more than sixty percent in most cases without negatively affecting accuracy.  Also, decision and model trees built from the pre-processed data are often significantly smaller.
867&lt;/p&gt;&lt;/div&gt;
868
869
870
871&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
872
873&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
874
875&lt;/p&gt;&lt;/div&gt;
876
877
878
879&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
880
881&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
882
883&lt;/p&gt;&lt;/div&gt;
884
885
886
887&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
888
889&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
89099/5
891&lt;/p&gt;&lt;/div&gt;
892
893
894
895&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
896
897&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
898Browsing tree structures
899&lt;/p&gt;&lt;/div&gt;
900
901
902
903&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
904
905&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
906Mark Apperley, Robert Spence, Stephen Hodge, Michael Chester
907&lt;/p&gt;&lt;/div&gt;
908
909
910
911&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
912
913&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
914Graphic representations of tree structures are notoriously difficult to create, display, and interpret, particularly when the volume of information they contain, and hence the number of nodes, is large.  The problem of interactively browsing information held in tree structures is examined, and the implementation of an innovative tree browser described.  This browser is based on distortion-oriented display techniques and intuitive direct manipulation interaction.  The tree layout is automatically generated, but the location and extent of detail shown is controlled by the user.  It is suggested that these techniques could be extended to the browsing of more general networks.
915&lt;/p&gt;&lt;/div&gt;
916
917
918
919&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
920
921&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
922
923&lt;/p&gt;&lt;/div&gt;
924
925
926
927&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
928
929&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
93099/6
931&lt;/p&gt;&lt;/div&gt;
932
933
934
935&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
936
937&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
938Facilitating multiple copy/past operations
939&lt;/p&gt;&lt;/div&gt;
940
941
942
943&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
944
945&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
946Mark Apperley, Jay Baker, Dale Fletcher, Bill Rogers
947&lt;/p&gt;&lt;/div&gt;
948
949
950
951&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
952
953&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
954Copy and paste, or cut and paste, using a clipboard or paste buffer has long been the principle facility provided to users for transferring data between and within GUI applications.  We argue that this mechanism can be clumsy in circumstances where several pieces of information must be moved systematically.  In two situations - extraction of data fields from unstructured data found in a directed search process, and reorganisation of computer program source text - we present alternative, more natural, user interface facilities to make the task less onerous, and to provide improved visual feedback during the operation.
955&lt;/p&gt;&lt;/div&gt;
956
957
958
959&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
960
961&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
962
963&lt;/p&gt;&lt;/div&gt;
964
965
966
967&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
968
969&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
970For the data extraction task we introduce the Stretchable Selection Tool, a semi-transparent overlay augmenting the mouse pointer to automate paste operations and provide information to prompt the user.  We describe a prototype implementation that functions in a collaborative software environment, allowing users to cooperate on a multiple copy/paste operation.  For text reorganisation, we present an extension to Emacs, providing similar functionality, but without the collaborative features.
971&lt;/p&gt;&lt;/div&gt;
972
973
974
975&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
976
977&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
978
979&lt;/p&gt;&lt;/div&gt;
980
981
982
983&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
984
985&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
98699/7
987&lt;/p&gt;&lt;/div&gt;
988
989
990
991&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
992
993&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
994Automating iterative tasks with programming by demonstration: a user evaluation
995&lt;/p&gt;&lt;/div&gt;
996
997
998
999&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1000
1001&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1002Gordon W. Paynter, Ian H. Witten
1003&lt;/p&gt;&lt;/div&gt;
1004
1005
1006
1007&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1008
1009&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1010Computer users often face iterative tasks that cannot be automated using the tools and aggregation techniques provided by their application program: they end up performing the iteration by hand, repeating user interface actions over and over again.  We have implemented an agent, called Familiar, that can be taught to perform iterative tasks using programming by demonstration (PBD).  Unlike other PBD systems, it is domain independent and works with unmodified, widely-used, applications in a popular operating system.  In a formal evaluation, we found that users quickly learned to use the agent to automate iterative tasks.  Generally, the participants preferred to use multiple selection where possible, but could and did use PBD in situations involving iteration over many commands, or when other techniques were unavailable.
1011&lt;/p&gt;&lt;/div&gt;
1012
1013
1014
1015&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1016
1017&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1018
1019&lt;/p&gt;&lt;/div&gt;
1020
1021
1022
1023&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1024
1025&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
102699/8
1027&lt;/p&gt;&lt;/div&gt;
1028
1029
1030
1031&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1032
1033&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1034A survey of software requirements specification practices in the New Zealand software industry
1035&lt;/p&gt;&lt;/div&gt;
1036
1037
1038
1039&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1040
1041&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1042Lindsay Groves, Ray Nickson, Greg Reeve, Steve Reeves, Mark Utting
1043&lt;/p&gt;&lt;/div&gt;
1044
1045
1046
1047&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1048
1049&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1050We report on the software development techniques used in the New Zealand software industry, paying particular attention to requirements gathering.  We surveyed a selection of software companies with a general questionnaire and then conducted in-depth interviews with four companies.  Our results show a wide variety in the kinds of companies undertaking software development, employing a wide range of software development techniques.  Although our data are not sufficiently detailed to draw statistically significant conclusions, it appears that larger software development groups typically have more well-defined software development processes, spend proportionally more time on requirements gathering, and follow more rigorous testing regimes.
1051&lt;/p&gt;&lt;/div&gt;
1052
1053
1054
1055&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1056
1057&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1058
1059&lt;/p&gt;&lt;/div&gt;
1060
1061
1062
1063&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1064
1065&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
106699/9
1067&lt;/p&gt;&lt;/div&gt;
1068
1069
1070
1071&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1072
1073&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1074The LRU*WWW proxy cache document replacement algorithm
1075&lt;/p&gt;&lt;/div&gt;
1076
1077
1078
1079&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1080
1081&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1082Chung-yi Chang, Tony McGregor, Geoffrey Holmes
1083&lt;/p&gt;&lt;/div&gt;
1084
1085
1086
1087&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1088
1089&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1090Obtaining good performance from WWW proxy caches is critically dependent on the document replacement policy used by the proxy.  This paper validates the work of other authors by reproducing their studies of proxy cache document replacement algorithms.  From this basis a cross-trace study is mounted.  This demonstrates that the performance of most document replacement algorithms is dependent on the type of workload that they are presented with.  Finally we propose a new algorithm, LRU*, that consistently performs well across all our traces.
1091&lt;/p&gt;&lt;/div&gt;
1092
1093
1094
1095&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1096
1097&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1098
1099&lt;/p&gt;&lt;/div&gt;
1100
1101
1102
1103&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1104
1105&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
110699/10
1107&lt;/p&gt;&lt;/div&gt;
1108
1109
1110
1111&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1112
1113&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1114Reduced-error pruning with significance tests
1115&lt;/p&gt;&lt;/div&gt;
1116
1117
1118
1119&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1120
1121&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1122Eibe Frank, Ian H. Witten
1123&lt;/p&gt;&lt;/div&gt;
1124
1125
1126
1127&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1128
1129&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1130When building classification models, it is common practice to prune them to counter spurious effects of the training data: this often improves performance and reduces model size.  &amp;quot;Reduced-error pruning&amp;quot; is a fast pruning procedure for decision trees that is known to produce small and accurate trees.  Apart from the data from which the tree is grown, it uses an independent &amp;quot;pruning&amp;quot; set, and pruning decisions are based on the model's error rate on this fresh data.  Recently it has been observed that reduced-error pruning overfits the pruning data, producing unnecessarily large decision trees.  This paper investigates whether standard statistical significance tests can be used to counter this phenomenon.
1131&lt;/p&gt;&lt;/div&gt;
1132
1133
1134
1135&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1136
1137&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1138
1139&lt;/p&gt;&lt;/div&gt;
1140
1141
1142
1143&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1144
1145&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1146The problem of overfitting to the pruning set highlights the need for significance testing.  We investigate two classes of test, &amp;quot;parametric&amp;quot; and &amp;quot;non-parametric.&amp;quot;  The standard chi-squared statistic can be used both in a parametric test and as the basis for a non-parametric permutation test.  In both cases it is necessary to select the significance level at which pruning is applied.  We show empirically that both versions of the chi-squared test perform equally well if their significance levels are adjusted appropriately.  Using a collection of standard datasets, we show that significance testing improves on standard reduced error pruning if the significance level is tailored to the particular dataset at hand using cross-validation, yielding consistently smaller trees that perform at least as well and sometimes better.
1147&lt;/p&gt;&lt;/div&gt;
1148
1149
1150
1151&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1152
1153&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1154
1155&lt;/p&gt;&lt;/div&gt;
1156
1157
1158
1159&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1160
1161&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
116299/11
1163&lt;/p&gt;&lt;/div&gt;
1164
1165
1166
1167&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1168
1169&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1170Weka: Practical machine learning tools and techniques with Java implementations
1171&lt;/p&gt;&lt;/div&gt;
1172
1173
1174
1175&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1176
1177&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1178Ian H. Witten, Eibe Frank, Len Trigg, Mark Hall, Geoffrey Holmes, Sally Jo Cunningham
1179&lt;/p&gt;&lt;/div&gt;
1180
1181
1182
1183&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1184
1185&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1186The Waikato Environment for Knowledge Analysis (Weka) is a comprehensive suite of Java class libraries that implement many state-of-the-art machine learning and data mining algorithms.  Weka is freely available on the World-Wide Web and accompanies a new text on data mining [1] which documents and fully explains all the algorithms it contains.  Applications written using the Weka class libraries can be run on any computer with a Web browsing capability; this allows users to apply machine learning techniques to their own data regardless of computer platform.
1187&lt;/p&gt;&lt;/div&gt;
1188
1189
1190
1191&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1192
1193&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1194
1195&lt;/p&gt;&lt;/div&gt;
1196
1197
1198
1199&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1200
1201&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
120299/12
1203&lt;/p&gt;&lt;/div&gt;
1204
1205
1206
1207&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1208
1209&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1210Pace Regression
1211&lt;/p&gt;&lt;/div&gt;
1212
1213
1214
1215&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1216
1217&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1218Yong Wang, Ian H. Witten
1219&lt;/p&gt;&lt;/div&gt;
1220
1221
1222
1223&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1224
1225&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1226This paper articulates a new method of linear regression, &amp;ldquo;pace regression&amp;rdquo;, that addresses many drawbacks of standard regression reported in the literature&amp;mdash;particularly the subset selection problem.  Pace regression improves on classical ordinary least squares (OLS) regression by evaluating the effect of each variable and using a clustering analysis to improve the statistical basis for estimating their contribution to the overall regression.  As well as outperforming OLS, it also outperforms&amp;mdash;in a remarkably general sense&amp;mdash;other linear modeling techniques in the literature, including subset selection procedures, which seek a reduction in dimensionality that falls out as a natural byproduct of pace regression.  The paper defines six procedures that share the fundamental idea of pace regression, all of which are theoretically justified in terms of asymptotic performance.  Experiments confirm the performance improvement over other techniques.
1227&lt;/p&gt;&lt;/div&gt;
1228
1229
1230
1231&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1232
1233&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1234
1235&lt;/p&gt;&lt;/div&gt;
1236
1237
1238
1239&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1240
1241&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
124299/13
1243&lt;/p&gt;&lt;/div&gt;
1244
1245
1246
1247&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1248
1249&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1250A compression-based algorithm for Chinese word segmentation
1251&lt;/p&gt;&lt;/div&gt;
1252
1253
1254
1255&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1256
1257&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1258W.J. Teahan, Yingying Wen, Rodger McNab, Ian H. Witten
1259&lt;/p&gt;&lt;/div&gt;
1260
1261
1262
1263&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1264
1265&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1266The Chinese language is written without using spaces or other word delimiters.  Although a text may be thought of as a corresponding sequence of words, there is considerable ambiguity in the placement of boundaries.  Interpreting a text as a sequence of words is beneficial for some information retrieval and storage tasks: for example, full-text search, word-based compression, and keyphrase extraction.
1267&lt;/p&gt;&lt;/div&gt;
1268
1269
1270
1271&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1272
1273&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1274
1275&lt;/p&gt;&lt;/div&gt;
1276
1277
1278
1279&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1280
1281&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1282We describe a scheme that infers appropriate positions for word boundaries using an adaptive language model that is standard in text compression.  It is trained on a corpus of pre-segmented text, and when applied to new text, interpolates word boundaries so as to maximize the compression obtained.  This simple and general method performs well with respect to specialized schemes for Chinese language segmentation.
1283&lt;/p&gt;&lt;/div&gt;
1284
1285
1286
1287&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1288
1289&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1290
1291&lt;/p&gt;&lt;/div&gt;
1292
1293
1294
1295&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1296
1297&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
129899/14
1299&lt;/p&gt;&lt;/div&gt;
1300
1301
1302
1303&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1304
1305&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1306Clustering with finite data from semi-parametric mixture distributions
1307&lt;/p&gt;&lt;/div&gt;
1308
1309
1310
1311&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1312
1313&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1314Yong Wang, Ian H. Witten
1315&lt;/p&gt;&lt;/div&gt;
1316
1317
1318
1319&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1320
1321&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1322Existing clustering methods for the semi-parametric mixture distribution perform well as the volume of data increases.  However, they all suffer from a serious drawback in finite-data situations: small outlying groups of data points can be completely ignored in the clusters that are produced, no matter how far away they lie from the major clusters.  This can result in unbounded loss if the loss function is sensitive to the distance between clusters.
1323&lt;/p&gt;&lt;/div&gt;
1324
1325
1326
1327&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1328
1329&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1330This paper proposes a new distance-based clustering method that overcomes the problem by avoiding global constraints.  Experimental results illustrate its superiority to existing methods when small clusters are present in finite data sets; they also suggest that it is more accurate and stable than other methods even when there are no small clusters.
1331&lt;/p&gt;&lt;/div&gt;
1332
1333
1334
1335&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1336
1337&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1338
1339&lt;/p&gt;&lt;/div&gt;
1340
1341
1342
1343&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1344
1345&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
134699/15
1347&lt;/p&gt;&lt;/div&gt;
1348
1349
1350
1351&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1352
1353&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1354
1355&lt;/p&gt;&lt;/div&gt;
1356
1357
1358
1359&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1360
1361&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1362
1363&lt;/p&gt;&lt;/div&gt;
1364
1365
1366
1367&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1368
1369&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
137099/16
1371&lt;/p&gt;&lt;/div&gt;
1372
1373
1374
1375&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1376
1377&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1378The Niupepa Collection:  Opening the blinds on a window to the past
1379&lt;/p&gt;&lt;/div&gt;
1380
1381
1382
1383&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1384
1385&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1386Te Taka Keegan, Sally Jo Cunningham, Mark Apperley
1387&lt;/p&gt;&lt;/div&gt;
1388
1389
1390
1391&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1392
1393&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1394This paper describes the building of a digital library collection of historic newspapers.  The newspapers (&lt;i&gt;Niupepa&lt;/i&gt; in Maori), which were published in New Zealand during the period 1842 to 1933, form a unique historical record of the Maori language, and of events from an historical perspective.  Images of these newspapers have been converted to digital form, electronic text extracted from these, and the collection is now being made available over the Internet as a part of the New Zealand Digital Library (NZDL) project at the University of Waikato.
1395&lt;/p&gt;&lt;/div&gt;
1396
1397
1398
1399&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1400
1401&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1402
1403&lt;/p&gt;&lt;/div&gt;
1404
1405
1406
1407&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1408
1409&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1410
1411&lt;/p&gt;&lt;/div&gt;
1412
1413
1414
1415&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1416
1417&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1418
1419&lt;/p&gt;&lt;/div&gt;
1420
1421
1422
1423&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1424
1425&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
142698/1
1427&lt;/p&gt;&lt;/div&gt;
1428
1429
1430
1431&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1432
1433&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1434Boosting trees for cost-sensitive classifications
1435&lt;/p&gt;&lt;/div&gt;
1436
1437
1438
1439&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1440
1441&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1442Kai Ming Ting, Zijian Zheng
1443&lt;/p&gt;&lt;/div&gt;
1444
1445
1446
1447&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1448
1449&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1450This paper explores two boosting techniques for cost-sensitive tree classification in the situation where misclassification costs change very often.  Ideally, one would like to have only one induction, and use the induced model for different misclassification costs.  Thus, it demands robustness of the induced model against cost changes.  Combining multiple trees gives robust predictions against this change.  We demonstrate that ordinary boosting combined with the minimum expected cost criterion to select the prediction class is a good solution under this situation.  We also introduce a variant of the ordinary boosting procedure which utilizes the cost information during training.  We show that the proposed technique performs better than the ordinary boosting in terms of misclassification cost.  However, this technique requires to induce a set of new trees every time the cost changes.  Our empirical investigation also reveals some interesting behavior of boosting decision trees for cost-sensitive classification.
1451&lt;/p&gt;&lt;/div&gt;
1452
1453
1454
1455&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1456
1457&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1458
1459&lt;/p&gt;&lt;/div&gt;
1460
1461
1462
1463&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1464
1465&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1466
1467&lt;/p&gt;&lt;/div&gt;
1468
1469
1470
1471&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1472
1473&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
147498/2
1475&lt;/p&gt;&lt;/div&gt;
1476
1477
1478
1479&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1480
1481&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1482Generating accurate rule sets without global optimization
1483&lt;/p&gt;&lt;/div&gt;
1484
1485
1486
1487&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1488
1489&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1490Eibe Frank, Ian H. Witten
1491&lt;/p&gt;&lt;/div&gt;
1492
1493
1494
1495&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1496
1497&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1498The two dominant schemes for rule-learning, C4.5 and RIPPER, both operate in two stages.  First they induce an initial rule set and then they refine it using a rather complex optimization stage that discards (C4.5) or adjusts (RIPPER) individual rules to make them work better together.  In contrast, this paper shows how good rule sets can be learned one rule at a time, without any need for global optimization.  We present an algorithm for inferring rules by repeatedly generating partial decision trees, thus combining the two major paradigms for rule generation-creating rules from decision trees and the separate-and-conquer rule-learning technique.  The algorithm is straightforward and elegant: despite this, experiments on standard datasets show that it produces rule sets that are as accurate as and of similar size to those generated by C4.5, and more accurate than RIPPER's.  Moreover, it operates efficiently, and because it avoids postprocessing, does not suffer the extremely slow performance on pathological example sets for which the C4.5 method has been criticized.
1499&lt;/p&gt;&lt;/div&gt;
1500
1501
1502
1503&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1504
1505&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1506
1507&lt;/p&gt;&lt;/div&gt;
1508
1509
1510
1511&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1512
1513&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1514
1515&lt;/p&gt;&lt;/div&gt;
1516
1517
1518
1519&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1520
1521&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
152298/3
1523&lt;/p&gt;&lt;/div&gt;
1524
1525
1526
1527&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1528
1529&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1530VQuery: a graphical user interface for Boolean query Specification and dynamic result preview
1531&lt;/p&gt;&lt;/div&gt;
1532
1533
1534
1535&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1536
1537&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1538Steve Jones
1539&lt;/p&gt;&lt;/div&gt;
1540
1541
1542
1543&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1544
1545&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1546Textual query languages based on Boolean logic are common amongst the search facilities of on-line information repositories.  However, there is evidence to suggest that the syntactic and semantic demands of such languages lead to user errors and adversely affect the time that it takes users to form queries.  Additionally, users are faced with user interfaces to these repositories which are unresponsive and uninformative, and consequently fail to support effective query refinement.  We suggest that graphical query languages, particularly Venn-like diagrams, provide a natural medium for Boolean query specification which overcomes the problems of textual query languages.  Also, dynamic result previews can be seamlessly integrated with graphical query specification to increase the effectiveness of query refinements.  We describe VQuery, a query interface to the New Zealand Digital Library which exploits querying by Venn diagrams and integrated query result previews.
1547&lt;/p&gt;&lt;/div&gt;
1548
1549
1550
1551&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1552
1553&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1554
1555&lt;/p&gt;&lt;/div&gt;
1556
1557
1558
1559&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1560
1561&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1562
1563&lt;/p&gt;&lt;/div&gt;
1564
1565
1566
1567&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1568
1569&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
157098/4
1571&lt;/p&gt;&lt;/div&gt;
1572
1573
1574
1575&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1576
1577&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1578Revising &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;: semantics and logic
1579&lt;/p&gt;&lt;/div&gt;
1580
1581
1582
1583&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1584
1585&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1586Martin C. Henson, Steve Reeves
1587&lt;/p&gt;&lt;/div&gt;
1588
1589
1590
1591&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1592
1593&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1594We introduce a simple specification logic &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;c comprising a logic and semantics (in &amp;lt;I&amp;gt;ZF&amp;lt;/I&amp;gt; set theory).  We then provide an interpretation for (a rational reconstruction of) the specification language &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt; within &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;c.  As a result we obtain a sound logic for &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;, including the schema calculus.  A consequence of our formalisation is a critique of a number of concepts used in &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;.  We demonstrate that the complications and confusions which these concepts introduce can be avoided without compromising expressibility.
1595&lt;/p&gt;&lt;/div&gt;
1596
1597
1598
1599&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1600
1601&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1602
1603&lt;/p&gt;&lt;/div&gt;
1604
1605
1606
1607&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1608
1609&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1610
1611&lt;/p&gt;&lt;/div&gt;
1612
1613
1614
1615&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1616
1617&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
161898/5
1619&lt;/p&gt;&lt;/div&gt;
1620
1621
1622
1623&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1624
1625&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1626A logic for the schema calculus
1627&lt;/p&gt;&lt;/div&gt;
1628
1629
1630
1631&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1632
1633&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1634Martin C. Henson, Steve Reeves
1635&lt;/p&gt;&lt;/div&gt;
1636
1637
1638
1639&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1640
1641&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1642In this paper we introduce and investigate a logic for the schema calculus of &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;.  The schema calculus is arguably the reason for &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;'s popularity but so far no true calculus (a sound system of rules for reasoning about schema expressions) has been given.  Presentations thus far have either failed to provide a calculus (e.g. the draft standard [3]) or have fallen back on informal descriptions at a syntactic level (most text books e.g. [7[).  Once the calculus is established we introduce a derived equational logic which enables us to formalise properly the informal notations of schema expression equality to be found in the literature.
1643&lt;/p&gt;&lt;/div&gt;
1644
1645
1646
1647&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1648
1649&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1650
1651&lt;/p&gt;&lt;/div&gt;
1652
1653
1654
1655&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1656
1657&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1658
1659&lt;/p&gt;&lt;/div&gt;
1660
1661
1662
1663&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1664
1665&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
166698/6
1667&lt;/p&gt;&lt;/div&gt;
1668
1669
1670
1671&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1672
1673&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1674New foundations for &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt;
1675&lt;/p&gt;&lt;/div&gt;
1676
1677
1678
1679&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1680
1681&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1682Martin C. Henson, Steve Reeves
1683&lt;/p&gt;&lt;/div&gt;
1684
1685
1686
1687&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1688
1689&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1690We provide a constructive and intensional interpretation for the specification language &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt; in a theory of operations and kinds &amp;lt;I&amp;gt;T&amp;lt;/I&amp;gt;.  The motivation is to facilitate the development of an integrated approach to program construction.  We illustrate the new foundations for &amp;lt;I&amp;gt;Z&amp;lt;/I&amp;gt; with examples.
1691&lt;/p&gt;&lt;/div&gt;
1692
1693
1694
1695&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1696
1697&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1698
1699&lt;/p&gt;&lt;/div&gt;
1700
1701
1702
1703&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1704
1705&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1706
1707&lt;/p&gt;&lt;/div&gt;
1708
1709
1710
1711&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1712
1713&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
171498/7
1715&lt;/p&gt;&lt;/div&gt;
1716
1717
1718
1719&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1720
1721&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1722Predicting apple bruising relationships using machine learning
1723&lt;/p&gt;&lt;/div&gt;
1724
1725
1726
1727&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1728
1729&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1730G. Holmes, S.J. Cunningham, B.T. Dela Rue, A.F. Bollen
1731&lt;/p&gt;&lt;/div&gt;
1732
1733
1734
1735&lt;p&gt;&lt;div name=&quot;Body Text&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1736
1737&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1738Many models have been used to describe the influence of internal or external factors on apple bruising.  Few of these have addressed the application of derived relationships to the evaluation of commercial operations.  From an industry perspective, a model must enable fruit to be rejected on the basis of a commercially significant bruise and must also accurately quantify the effects of various combinations of input features (such as cultivar, maturity, size, and so on) on bruise prediction.  Input features must in turn have characteristics which are measurable commercially; for example, the measure of force should be impact energy rather than energy absorbed.  Further, as the commercial criteria for acceptable damage levels change, the model should be versatile enough to regenerate new bruise thresholds from existing data.
1739&lt;/p&gt;&lt;/div&gt;
1740
1741
1742
1743&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1744
1745&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1746
1747&lt;/p&gt;&lt;/div&gt;
1748
1749
1750
1751&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1752
1753&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1754Machine learning is a burgeoning technology with a vast range of potential applications particularly in agriculture where large amounts of data can be readily collected [1].  The main advantage of using a machine learning method in an application is that the models built for prediction can be viewed and understood by the owner of the data who is in a position to determine the usefulness of the model, an essential component in a commercial environment.
1755&lt;/p&gt;&lt;/div&gt;
1756
1757
1758
1759&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1760
1761&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1762
1763&lt;/p&gt;&lt;/div&gt;
1764
1765
1766
1767&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1768
1769&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1770
1771&lt;/p&gt;&lt;/div&gt;
1772
1773
1774
1775&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1776
1777&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
177898/8
1779&lt;/p&gt;&lt;/div&gt;
1780
1781
1782
1783&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1784
1785&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1786An evaluation of passage-level indexing strategies for a technical report archive
1787&lt;/p&gt;&lt;/div&gt;
1788
1789
1790
1791&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1792
1793&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1794Michael Williams
1795&lt;/p&gt;&lt;/div&gt;
1796
1797
1798
1799&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1800
1801&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1802Past research has shown that using evidence from document passages rather than complete documents is an effective way of improving the precision of full-text database searches.  However, passage-level indexing has yet to be widely adopted for commercial or online databases.
1803&lt;/p&gt;&lt;/div&gt;
1804
1805
1806
1807&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1808
1809&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1810
1811&lt;/p&gt;&lt;/div&gt;
1812
1813
1814
1815&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1816
1817&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1818This paper reports on experiments designed to test the efficacy of passage-level indexing with a particular collection of a full-text online database, the New Zealand Digital Library.  Discourse passages and word-window passages are used for the indexing process.  Both ranked and Boolean searching are used to test the resulting indexes.
1819&lt;/p&gt;&lt;/div&gt;
1820
1821
1822
1823&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1824
1825&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1826
1827&lt;/p&gt;&lt;/div&gt;
1828
1829
1830
1831&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1832
1833&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1834Overlapping window passages are shown to offer the best retrieval performance with both ranked and Boolean queries.  Modifications may be necessary to the term weighting methodology in order to ensure optimal ranked query performance.
1835&lt;/p&gt;&lt;/div&gt;
1836
1837
1838
1839&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1840
1841&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1842
1843&lt;/p&gt;&lt;/div&gt;
1844
1845
1846
1847&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1848
1849&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1850
1851&lt;/p&gt;&lt;/div&gt;
1852
1853
1854
1855&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1856
1857&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
185898/9
1859&lt;/p&gt;&lt;/div&gt;
1860
1861
1862
1863&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1864
1865&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1866Managing multiple collections, multiple languages, and multiple media in a distributed digital library
1867&lt;/p&gt;&lt;/div&gt;
1868
1869
1870
1871&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1872
1873&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1874Ian H. Witten, Rodger McNab, Steve Jones, Sally Jo Cunningham, David Bainbridge, Mark Apperley
1875&lt;/p&gt;&lt;/div&gt;
1876
1877
1878
1879&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1880
1881&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1882Managing the organizational and software complexity of a comprehensive digital library presents a significant challenge.  Different library collections each have their own distinctive features.  Different presentation languages have structural implications such as left-to-right writing order and text-only interfaces for the visually impaired.  Different media involve different file formats, and-more importantly-radically different search strategies are required for non-textual media.  In a distributed library, new collections can appear asynchronously on servers in different parts of the world.  And as searching interfaces mature from the command-line era exemplified by current Web search engines into the age of reactive visual interfaces, experimental new interfaces must be developed, supported, and tested.  This paper describes our experience, gained from operating a substantial digital library service over several years, in solving these problems by designing an appropriate software architecture.
1883&lt;/p&gt;&lt;/div&gt;
1884
1885
1886
1887&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1888
1889&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1890
1891&lt;/p&gt;&lt;/div&gt;
1892
1893
1894
1895&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1896
1897&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1898
1899&lt;/p&gt;&lt;/div&gt;
1900
1901
1902
1903&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1904
1905&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
190698/10
1907&lt;/p&gt;&lt;/div&gt;
1908
1909
1910
1911&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1912
1913&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1914Experiences with a weighted decision tree learner
1915&lt;/p&gt;&lt;/div&gt;
1916
1917
1918
1919&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1920
1921&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1922John G. Cleary, Leonard E. Trigg
1923&lt;/p&gt;&lt;/div&gt;
1924
1925
1926
1927&lt;p&gt;&lt;div name=&quot;Body Text&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1928
1929&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1930Machine learning algorithms for inferring decision trees typically choose a single &amp;ldquo;best&amp;rdquo; tree to describe the training data.  Recent research has shown that classification performance can be significantly improved by voting predictions of multiple, independently produced decision trees.  This paper describes an algorithm, OB1, that makes a weighted sum over many possible models.  We describe one instance of OB1, that includes &amp;lt;I&amp;gt;all&amp;lt;/I&amp;gt; possible decision trees as well as naïve Bayesian models.  OB1 is compared with a number of other decision tree and instance based learning alogrithms on some of the data sets from the UCI repository.  Both an information gain and an accuracy measure are used for the comparison.  On the information gain measure OB1 performs significantly better than all the other algorithms.  On the accuracy measure it is significantly better than all the algorithms except naïve Bayes which performs comparably to OB1.
1931&lt;/p&gt;&lt;/div&gt;
1932
1933
1934
1935&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1936
1937&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1938
1939&lt;/p&gt;&lt;/div&gt;
1940
1941
1942
1943&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1944
1945&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1946
1947&lt;/p&gt;&lt;/div&gt;
1948
1949
1950
1951&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1952
1953&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
195498/11
1955&lt;/p&gt;&lt;/div&gt;
1956
1957
1958
1959&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1960
1961&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1962An entropy gain measure of numeric prediction performance
1963&lt;/p&gt;&lt;/div&gt;
1964
1965
1966
1967&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1968
1969&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1970Leonard Trigg
1971&lt;/p&gt;&lt;/div&gt;
1972
1973
1974
1975&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1976
1977&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1978Categorical classifier performance is typically evaluated with respect to error rate, expressed as a percentage of test instances that were not correctly classified.  When a classifier produces multiple classifications for a test instance, the prediction is counted as incorrect (even if the correct class was one of the predictions).  Although commonly used in the literature, error rate is a coarse measure of classifier performance, as it is based only on a single prediction offered for a test instance.  Since many classifiers can produce a class distribution as a prediction, we should use this to provide a better measure of how much information the classifier is extracting from the domain.
1979&lt;/p&gt;&lt;/div&gt;
1980
1981
1982
1983&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1984
1985&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1986
1987&lt;/p&gt;&lt;/div&gt;
1988
1989
1990
1991&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
1992
1993&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
1994Numeric classifiers are a relatively new development in machine learning, and as such there is no single performance measure that has become standard.  Typically these machine learning schemes predict a single real number for each test instance, and the error between the predicted and actual value is used to calculate a myriad of performance measures such as correlation coefficient, root mean squared error, mean absolute error, relative absolute error, and root relative squared error.  With so many performance measures it is difficult to establish an overall performance evaluation.
1995&lt;/p&gt;&lt;/div&gt;
1996
1997
1998
1999&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2000
2001&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2002
2003&lt;/p&gt;&lt;/div&gt;
2004
2005
2006
2007&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2008
2009&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2010The next section describes a performance measure for machine learning schemes that attempts to overcome the problems with current measures.  In addition, the same evaluation measure is used for categorical and numeric classifier.
2011&lt;/p&gt;&lt;/div&gt;
2012
2013
2014
2015&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2016
2017&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2018
2019&lt;/p&gt;&lt;/div&gt;
2020
2021
2022
2023&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2024
2025&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2026
2027&lt;/p&gt;&lt;/div&gt;
2028
2029
2030
2031&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2032
2033&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2034
2035&lt;/p&gt;&lt;/div&gt;
2036
2037
2038
2039&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2040
2041&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
204298/12
2043&lt;/p&gt;&lt;/div&gt;
2044
2045
2046
2047&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2048
2049&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2050Proceedings of CBISE '98 CaiSE*98 Workshop on Component Based Information Systems Engineering
2051&lt;/p&gt;&lt;/div&gt;
2052
2053
2054
2055&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2056
2057&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2058Edited by John Grundy
2059&lt;/p&gt;&lt;/div&gt;
2060
2061
2062
2063&lt;p&gt;&lt;div name=&quot;Body Text&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2064
2065&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2066Component-based information systems development is an area of research and practice of increasing importance.  Information Systems developers have realised that traditional approaches to IS engineering produce monolithic, difficult to maintain, difficult to reuse systems.  In contrast, the use of software components, which embody data, functionality and well-specified and understood interfaces, makes interoperable, distributed and highly reusable IS components feasible.  Component-based approaches to IS engineering can be used at strategic and organisational levels, to model business processes and whole IS architectures, in development methods which utilise component-based models during analysis and design, and in system implementation.  Reusable components can allow end users to compose and configure their own Information Systems, possibly from a range of suppliers, and to more tightly couple their organisational workflows with their IS support.
2067&lt;/p&gt;&lt;/div&gt;
2068
2069
2070
2071&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2072
2073&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2074
2075&lt;/p&gt;&lt;/div&gt;
2076
2077
2078
2079&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2080
2081&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2082This workshop proceedings contains a range of papers addressing one or more of the above issues relating to the use of component models for IS development.  All of these papers were refereed by at least two members of an international workshop committee comprising industry and academic researchers and users of component technologies.  Strategic uses of components are addressed in the first three papers, while the following three address uses of components for systems design and workflow management.  Systems development using components, and the provision of environments for component management are addressed in the following group of five papers.  The last three papers in this proceedings address component management and analysis techniques.
2083&lt;/p&gt;&lt;/div&gt;
2084
2085
2086
2087&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2088
2089&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2090
2091&lt;/p&gt;&lt;/div&gt;
2092
2093
2094
2095&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2096
2097&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2098All of these papers provide new insights into the many  varied uses of component technology for IS engineering.  I hope you find them as interesting and useful as I have when collating this proceedings and organising the workshop.
2099&lt;/p&gt;&lt;/div&gt;
2100
2101
2102
2103&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2104
2105&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2106
2107&lt;/p&gt;&lt;/div&gt;
2108
2109
2110
2111&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2112
2113&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2114
2115&lt;/p&gt;&lt;/div&gt;
2116
2117
2118
2119&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2120
2121&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
212298/13
2123&lt;/p&gt;&lt;/div&gt;
2124
2125
2126
2127&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2128
2129&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2130An analysis of usage of a digital library
2131&lt;/p&gt;&lt;/div&gt;
2132
2133
2134
2135&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2136
2137&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2138Steve Jones, Sally Jo Cunningham, Rodger McNab
2139&lt;/p&gt;&lt;/div&gt;
2140
2141
2142
2143&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2144
2145&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2146As experimental digital library testbeds gain wider acceptance and develop significant user bases, it becomes important to investigate the ways in which users interact with the systems in practice.  Transaction logs are one source of usage information, and the information on user behaviour can be culled from them both automatically (through calculation of summary statistics) and manually (by examining query strings for semantic clues on search motivations and searching strategy).  We conduct a transaction log analysis on user activity in the Computer Science Technical Reports Collection of the New Zealand Digital Library, and report insights gained and identify resulting search interface design issues.
2147&lt;/p&gt;&lt;/div&gt;
2148
2149
2150
2151&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2152
2153&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2154
2155&lt;/p&gt;&lt;/div&gt;
2156
2157
2158
2159&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2160
2161&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2162
2163&lt;/p&gt;&lt;/div&gt;
2164
2165
2166
2167&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2168
2169&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
217098/14
2171&lt;/p&gt;&lt;/div&gt;
2172
2173
2174
2175&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2176
2177&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2178Measuring ATM traffic: final report for New Zealand Telecom
2179&lt;/p&gt;&lt;/div&gt;
2180
2181
2182
2183&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2184
2185&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2186John Cleary, Ian Graham, Murray Pearson, Tony McGregor
2187&lt;/p&gt;&lt;/div&gt;
2188
2189
2190
2191&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2192
2193&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2194The report describes the development of a low-cost ATM monitoring system, hosted by a standard PC.  The monitor can be used remotely returning information on ATM traffic flows to a central site.  The monitor is interfaces to a GPS timing receiver, which provides an absolute time accuracy of better than 1 usec.  By monitoring the same traffic flow at different points in a network it is possible to measure cell delay and delay variation in real time, and with existing traffic.  The monitoring system characterises cells by a CRC calculated over the cell payload, thus special measurement cells are not required.  Delays in both local area and wide-area networks have been measured using this system.  It is possible to measure delay in a network that is not end-to-end ATM, as long as some cells remain identical at the entry and exit points.  Examples are given of traffic and delay measurements in both wide and local area network systems, including delays measured over the Internet from Canada to New Zealand.
2195&lt;/p&gt;&lt;/div&gt;
2196
2197
2198
2199&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2200
2201&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2202
2203&lt;/p&gt;&lt;/div&gt;
2204
2205
2206
2207&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2208
2209&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2210
2211&lt;/p&gt;&lt;/div&gt;
2212
2213
2214
2215&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2216
2217&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2218
2219&lt;/p&gt;&lt;/div&gt;
2220
2221
2222
2223&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2224
2225&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
222698/15
2227&lt;/p&gt;&lt;/div&gt;
2228
2229
2230
2231&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2232
2233&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2234Despite its simplicity, the naïve Bayes learning scheme performs well on most classification tasks, and is often significantly more accurate than more sophisticated methods.  Although the probability estimates that it produces can be inaccurate, it often assigns maximum probability to the correct class.  This suggests that its good performance might be restricted to situations where the output is categorical.  It is therefore interesting to see how it performs in domains where the predicted value is numeric, because in this case, predictions are more sensitive to inaccurate probability estimates.&amp;lt;P&amp;gt;
2235&lt;/p&gt;&lt;/div&gt;
2236
2237
2238
2239&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2240
2241&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2242
2243&lt;/p&gt;&lt;/div&gt;
2244
2245
2246
2247&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2248
2249&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2250This paper shows how to apply the naïve Bayes methodology to numeric prediction (i.e. regression) tasks, and compares it to linear regression, instance-based learning, and a method that produces &amp;ldquo;model trees&amp;rdquo;-decision trees with linear regression functions at the leaves.  Although we exhibit an artificial dataset for which naïve Bayes is the method of choice, on real-world datasets it is almost uniformly worse than model trees.  The comparison with linear regression depends on the error measure: for one measure naïve Bayes performs similarly, for another it is worse.  Compared to instance-based learning, it performs similarly with respect to both measures.  These results indicate that the simplistic statistical assumption that naïve Bayes makes is indeed more restrictive for regression than for classification.
2251&lt;/p&gt;&lt;/div&gt;
2252
2253
2254
2255&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2256
2257&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2258
2259&lt;/p&gt;&lt;/div&gt;
2260
2261
2262
2263&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2264
2265&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
226698/16
2267&lt;/p&gt;&lt;/div&gt;
2268
2269
2270
2271&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2272
2273&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2274Link as you type: using key phrases for automated dynamic link generation
2275&lt;/p&gt;&lt;/div&gt;
2276
2277
2278
2279&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2280
2281&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2282Steve Jones
2283&lt;/p&gt;&lt;/div&gt;
2284
2285
2286
2287&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2288
2289&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2290When documents are collected together from diverse sources they are unlikely to contain useful hypertext links to support browsing amongst them.  For large collections of thousands of documents it is prohibitively resource intensive to manually insert links into each document.  Users of such collections may wish to relate documents within them to text that they are themselves generating.  This process, often involving keyword searching, distracts from the authoring process and results in material related to query terms but not necessarily to the author's document.  Query terms that are effective in one collection might not be so in another.  We have developed Phrasier, a system that integrates authoring (of text and hyperlinks), browsing, querying and reading in support of information retrieval activities.  Phrasier exploits key phrases which are automatically extracted from documents in a collection, and uses them as link anchors and to identify candidate destinations for hyperlinks.  This system suggests links into existing collections for purposes of authoring and retrieval of related information, creates links between documents in a collection and provides supportive document and link overviews.
2291&lt;/p&gt;&lt;/div&gt;
2292
2293
2294
2295&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2296
2297&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2298
2299&lt;/p&gt;&lt;/div&gt;
2300
2301
2302
2303&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2304
2305&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2306
2307&lt;/p&gt;&lt;/div&gt;
2308
2309
2310
2311&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2312
2313&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
231498/17
2315&lt;/p&gt;&lt;/div&gt;
2316
2317
2318
2319&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2320
2321&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2322Melody based tune retrieval over the World Wide Web
2323&lt;/p&gt;&lt;/div&gt;
2324
2325
2326
2327&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2328
2329&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2330David Bainbridge, Rodger J. McNab, Lloyd A. Smith
2331&lt;/p&gt;&lt;/div&gt;
2332
2333
2334
2335&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2336
2337&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2338In this paper we describe the steps taken to develop a Web-based version of an existing stand-alone, single-user digital library application for melodical searching of a collection of music.  For the three key components: input, searching, and output, we assess the suitability of various Web-based strategies that deal with the now distributed software architecture and explain the decisions we made.  The resulting melody indexing service, known as MELDEX, has been in operation for one year, and the feed-back we have received has been favorable.
2339&lt;/p&gt;&lt;/div&gt;
2340
2341
2342
2343&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2344
2345&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2346
2347&lt;/p&gt;&lt;/div&gt;
2348
2349
2350
2351&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2352
2353&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
235498/18
2355&lt;/p&gt;&lt;/div&gt;
2356
2357
2358
2359&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2360
2361&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2362Making oral history accessible over the World Wide Web
2363&lt;/p&gt;&lt;/div&gt;
2364
2365
2366
2367&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2368
2369&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2370David Bainbridge, Sally Jo Cunningham
2371&lt;/p&gt;&lt;/div&gt;
2372
2373
2374
2375&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2376
2377&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2378We describe a multimedia, WWW-based oral history collection constructed from off-the-shelf or publicly available software.  The source materials for the collection include audio tapes of interviews and summary transcripts of each interview, as well as photographs illustrating episodes mentioned in the tapes.  Sections of the transcripts are manually matched to associated segments of the tapes, and the tapes are digitized.  Users search a full-text retrieval system based on the text transcripts to retrieve relevant transcript sections and their associated audio recordings and photographs.  It is also possible to search for photos by matching text queries against text descriptions of the photos in the collection, where the located photos link back to their respective interview transcript and audio recordings.
2379&lt;/p&gt;&lt;/div&gt;
2380
2381
2382
2383&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2384
2385&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2386
2387&lt;/p&gt;&lt;/div&gt;
2388
2389
2390
2391&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2392
2393&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2394
2395&lt;/p&gt;&lt;/div&gt;
2396
2397
2398
2399&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2400
2401&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2402
2403&lt;/p&gt;&lt;/div&gt;
2404
2405
2406
2407&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2408
2409&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2410
2411&lt;/p&gt;&lt;/div&gt;
2412
2413
2414
2415&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2416
2417&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2418&lt;b&gt;1997&lt;/b&gt;
2419&lt;/p&gt;&lt;/div&gt;
2420
2421
2422
2423&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2424
2425&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2426
2427&lt;/p&gt;&lt;/div&gt;
2428
2429
2430
2431&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2432
2433&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
243497/1
2435&lt;/p&gt;&lt;/div&gt;
2436
2437
2438
2439&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2440
2441&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2442A dynamic and flexible representation of social relationships in CSCW
2443&lt;/p&gt;&lt;/div&gt;
2444
2445
2446
2447&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2448
2449&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2450Steve Jones, Steve Marsh
2451&lt;/p&gt;&lt;/div&gt;
2452
2453
2454
2455&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2456
2457&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2458CSCW system designers lack effective support in addressing the social issues and interpersonal relationships which are linked with the use of CSCW systems.  We present a formal description of trust to support CSCW system designers in considering the social aspects of group work, embedding those considerations in systems and analysing computer supported group processes.
2459&lt;/p&gt;&lt;/div&gt;
2460
2461
2462
2463&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2464
2465&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2466
2467&lt;/p&gt;&lt;/div&gt;
2468
2469
2470
2471&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2472
2473&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2474We argue that trust is a critical aspect in group work, and describe what we consider to be the building blocks of trust.  We then present a formal notation for the building blocks, their use in reasoning about social interactions and how they are amended over time.
2475&lt;/p&gt;&lt;/div&gt;
2476
2477
2478
2479&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2480
2481&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2482
2483&lt;/p&gt;&lt;/div&gt;
2484
2485
2486
2487&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2488
2489&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2490We then consider how the formalism may be used in practice, and present some insights from initial analysis of the behaviour of the formalism.  This is followed by a description of possible amendments and extensions to the formalism.  We conclude that it is possible to formalise a notion of trust and to model the formalisation by a computational mechanism.
2491&lt;/p&gt;&lt;/div&gt;
2492
2493
2494
2495&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2496
2497&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2498
2499&lt;/p&gt;&lt;/div&gt;
2500
2501
2502
2503&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2504
2505&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2506
2507&lt;/p&gt;&lt;/div&gt;
2508
2509
2510
2511&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2512
2513&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
251497/2
2515&lt;/p&gt;&lt;/div&gt;
2516
2517
2518
2519&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2520
2521&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2522Design issues for World Wide Web navigation visualisation tools
2523&lt;/p&gt;&lt;/div&gt;
2524
2525
2526
2527&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2528
2529&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2530Andy Cockburn, Steve Jones
2531&lt;/p&gt;&lt;/div&gt;
2532
2533
2534
2535&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2536
2537&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2538The World Wide Web (WWW) is a successful hypermedia information space used by millions of people, yet it suffers from many deficiencies and problems in support for navigation around its vast information space.  In this paper we identify the origins of these navigation problems, namely WWW browser design, WWW page design, and WWW page description languages.  Regardless of their origins, these problems are eventually represented to the user at the browser's user interface.  To help overcome these problems, many tools are being developed which allow users to visualise WWW subspaces.  We identify five key issues in the design and functionality of these visualisation systems: characteristics of the visual representation, the scope of the subspace representation, the mechanisms for generating the visualisation, the degree of browser independence, and the navigation support facilities.  We provide a critical review of the diverse range of WWW visualisation tools with respect to these issues.
2539&lt;/p&gt;&lt;/div&gt;
2540
2541
2542
2543&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2544
2545&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2546
2547&lt;/p&gt;&lt;/div&gt;
2548
2549
2550
2551&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2552
2553&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2554
2555&lt;/p&gt;&lt;/div&gt;
2556
2557
2558
2559&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2560
2561&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
256297/3
2563&lt;/p&gt;&lt;/div&gt;
2564
2565
2566
2567&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2568
2569&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2570Stacked generalization:  when does it work?
2571&lt;/p&gt;&lt;/div&gt;
2572
2573
2574
2575&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2576
2577&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2578Kai Ming Ting, Ian H. Witten
2579&lt;/p&gt;&lt;/div&gt;
2580
2581
2582
2583&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2584
2585&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2586Stacked generalization is a general method of using a high-level model to combine lower-level models to achieve greater predictive accuracy.  In this paper we address two crucial issues which have been considered to be a 'black art' in classification tasks ever since the introduction of stacked generalization in 1992 by Wolpert: the type of generalizer that is suitable to derive the higher-level model, and the kind of attributes that should be used as its input.
2587&lt;/p&gt;&lt;/div&gt;
2588
2589
2590
2591&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2592
2593&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2594
2595&lt;/p&gt;&lt;/div&gt;
2596
2597
2598
2599&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2600
2601&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2602We demonstrate the effectiveness of stacked generalization for combining three different types of learning algorithms, and also for combining models of the same type derived from a single learning algorithm in a multiple-data-batches scenario.  We also compare the performance of stacked generalization with published results arcing and bagging.
2603&lt;/p&gt;&lt;/div&gt;
2604
2605
2606
2607&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2608
2609&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2610
2611&lt;/p&gt;&lt;/div&gt;
2612
2613
2614
2615&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2616
2617&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2618
2619&lt;/p&gt;&lt;/div&gt;
2620
2621
2622
2623&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2624
2625&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
262697/4
2627&lt;/p&gt;&lt;/div&gt;
2628
2629
2630
2631&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2632
2633&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2634Browsing in digital libraries:  a phrase-based approach
2635&lt;/p&gt;&lt;/div&gt;
2636
2637
2638
2639&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2640
2641&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2642Craig Nevill-Manning, Ian H. Witten, Gordon W. Paynter
2643&lt;/p&gt;&lt;/div&gt;
2644
2645
2646
2647&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2648
2649&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2650A key question for digital libraries is this: how should one go about becoming familiar with a digital collection, as opposed to a physical one?  Digital collections generally present an appearance which is extremely opaque-a screen, typically a Web page, with no indication of what, or how much, lies beyond: whether a carefully-selected collection or a morass of worthless ephemera; whether half a dozen documents or many millions.  At least physical collections occupy physical space, present a physical appearance, and exhibit tangible physical organization.  When standing on the threshold of a large library one gains a sense of presence and permanence that reflects the care taken in building and maintaining the collection inside.  No-one could confuse it with a dung-heap!  Yet in the digital world the difference is not so palpable.
2651&lt;/p&gt;&lt;/div&gt;
2652
2653
2654
2655&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2656
2657&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2658
2659&lt;/p&gt;&lt;/div&gt;
2660
2661
2662
2663&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2664
2665&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2666
2667&lt;/p&gt;&lt;/div&gt;
2668
2669
2670
2671&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2672
2673&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2674
2675&lt;/p&gt;&lt;/div&gt;
2676
2677
2678
2679&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2680
2681&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
268297/5
2683&lt;/p&gt;&lt;/div&gt;
2684
2685
2686
2687&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2688
2689&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2690A graphical notation for the design of information visualisations
2691&lt;/p&gt;&lt;/div&gt;
2692
2693
2694
2695&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2696
2697&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2698Matthew C. Humphrey
2699&lt;/p&gt;&lt;/div&gt;
2700
2701
2702
2703&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2704
2705&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2706Visualisations are coherent, graphical expressions of complex information that enhance people's ability to communicate and reason about that information.  Yet despite the importance of visualisations in helping people to understand and solve a wide variety of problems, there is a dearth of formal tools and methods for discussing, describing and designing them.  Although simple visualisations, such as bar charts and scatterplots, are easily produced by modern interactive software, novel visualisations of multivariate, multirelational data must be expressed in a programming language.  The Relational Visualisation Notation is a new, graphical language for designing such highly expressive visualisations that does not use programming constructs.  Instead, the notation is based on relational algebra, which is widely used in database query languages, and it is supported by a suite of direct manipulation tools.  This article presents the notation and examines the designs of some interesting visualisations.
2707&lt;/p&gt;&lt;/div&gt;
2708
2709
2710
2711&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2712
2713&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2714
2715&lt;/p&gt;&lt;/div&gt;
2716
2717
2718
2719&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2720
2721&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2722
2723&lt;/p&gt;&lt;/div&gt;
2724
2725
2726
2727&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2728
2729&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2730
2731&lt;/p&gt;&lt;/div&gt;
2732
2733
2734
2735&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2736
2737&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
273897/6
2739&lt;/p&gt;&lt;/div&gt;
2740
2741
2742
2743&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2744
2745&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2746Applications of machine learning in information retrieval
2747&lt;/p&gt;&lt;/div&gt;
2748
2749
2750
2751&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2752
2753&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2754Sally Jo Cunningham, James Littin, Ian H. Witten
2755&lt;/p&gt;&lt;/div&gt;
2756
2757
2758
2759&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2760
2761&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2762Information retrieval systems provide access to collections of thousands, or millions, of documents, from which, by providing an appropriate description, users can recover any one.  Typically, users iteratively refine the descriptions they provide to satisfy their needs, and retrieval systems can utilize user feedback on selected documents to indicate the accuracy of the description at any stage.  The style of description required from the user, and the way it is employed to search the document database, are consequences of the indexing method used for the collection.  The index may take different forms, from storing keywords with links to individual documents, to clustering documents under related topics.
2763&lt;/p&gt;&lt;/div&gt;
2764
2765
2766
2767&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2768
2769&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2770
2771&lt;/p&gt;&lt;/div&gt;
2772
2773
2774
2775&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2776
2777&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2778
2779&lt;/p&gt;&lt;/div&gt;
2780
2781
2782
2783&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2784
2785&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2786
2787&lt;/p&gt;&lt;/div&gt;
2788
2789
2790
2791&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2792
2793&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
279497/7
2795&lt;/p&gt;&lt;/div&gt;
2796
2797
2798
2799&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2800
2801&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2802Computer concepts without computers:  a first course in computer science
2803&lt;/p&gt;&lt;/div&gt;
2804
2805
2806
2807&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2808
2809&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2810Geoffrey Holmes, Tony C. Smith, William J. Rogers
2811&lt;/p&gt;&lt;/div&gt;
2812
2813
2814
2815&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2816
2817&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2818While some institutions seek to make CS1 curricula more enjoyable by incorporating specialised educational software [1] or by setting more enjoyable programming assignments [2], we have joined the growing number of Computer Science departments that seek to improve the quality of the CS1 experience by focusing student attention away from the computer monitor [3,4].  Sophisticated computing concepts usually reserved for senior level courses are presented in a &amp;lt;I&amp;gt;popular science&amp;lt;/I&amp;gt; manner, and given equal time alongside the essential introductory programming material.  By exposing students to a broad range of specific computational problems we endeavour to make the introductory course more interesting and enjoyable, and instil in students a sense of vision for areas they might specialise in as computing majors.
2819&lt;/p&gt;&lt;/div&gt;
2820
2821
2822
2823&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2824
2825&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2826
2827&lt;/p&gt;&lt;/div&gt;
2828
2829
2830
2831&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2832
2833&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2834
2835&lt;/p&gt;&lt;/div&gt;
2836
2837
2838
2839&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2840
2841&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2842
2843&lt;/p&gt;&lt;/div&gt;
2844
2845
2846
2847&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2848
2849&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
285097/8
2851&lt;/p&gt;&lt;/div&gt;
2852
2853
2854
2855&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2856
2857&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2858A sight-singing tutor
2859&lt;/p&gt;&lt;/div&gt;
2860
2861
2862
2863&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2864
2865&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2866Lloyd A. Smith, Rodger J. McNab
2867&lt;/p&gt;&lt;/div&gt;
2868
2869
2870
2871&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2872
2873&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2874This paper describes a computer program designed to aid its users in learning to sight-sing.  Sight-singing-the ability to sing music from a score without prior study-is an important skill for musicians and holds a central place in most university music curricula.  Its importance to vocalists is obvious; it is also an important skill for instrumentalists and conductors because it develops the aural imagination necessary to judge how the music should sound, when played (Benward and Carr 1991).  Furthermore, it is an important skill for amateur musicians, who can save a great deal of rehearsal time through an ability to sing music at sight.
2875&lt;/p&gt;&lt;/div&gt;
2876
2877
2878
2879&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2880
2881&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2882
2883&lt;/p&gt;&lt;/div&gt;
2884
2885
2886
2887&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2888
2889&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2890
2891&lt;/p&gt;&lt;/div&gt;
2892
2893
2894
2895&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2896
2897&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
289897/9
2899&lt;/p&gt;&lt;/div&gt;
2900
2901
2902
2903&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2904
2905&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2906Stacking bagged and dagged models
2907&lt;/p&gt;&lt;/div&gt;
2908
2909
2910
2911&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2912
2913&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2914Kai Ming Ting, I.H. Witten
2915&lt;/p&gt;&lt;/div&gt;
2916
2917
2918
2919&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2920
2921&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2922In this paper, we investigate the method of &lt;i&gt;stacked generalization&lt;/i&gt; in combining models derived from different subsets of a training dataset by a single learning algorithm, as well as different algorithms.  The simplest way to combine predictions from competing models is majority vote, and the effect of the sampling regime used to generate training subsets has already been studied in this context-when bootstrap samples are used the method is called &lt;i&gt;bagging&lt;/i&gt;, and for disjoint samples we call it &lt;i&gt;dagging&lt;/i&gt;.  This paper extends these studies to stacked generalization, where a learning algorithm is employed to combine the models.  This yields new methods dubbed &lt;i&gt;bag-stacking&lt;/i&gt; and &lt;i&gt;dag-stacking&lt;/i&gt;.
2923&lt;/p&gt;&lt;/div&gt;
2924
2925
2926
2927&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2928
2929&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2930
2931&lt;/p&gt;&lt;/div&gt;
2932
2933
2934
2935&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2936
2937&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2938We demonstrate that bag-stacking and dag-stacking can be effective for classification tasks even when the training samples cover just a small fraction of the full dataset.  In contrast to earlier bagging results, we show that bagging and bag-stacking work for stable as well as unstable learning algorithms, as do dagging and dag-stacking.  We find that bag-stacking (dag-stacking) almost always has higher predictive accuracy than bagging (dagging), and we also show that bag-stacking models derived using two different algorithms is more effective than bagging.
2939&lt;/p&gt;&lt;/div&gt;
2940
2941
2942
2943&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2944
2945&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2946
2947&lt;/p&gt;&lt;/div&gt;
2948
2949
2950
2951&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2952
2953&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2954
2955&lt;/p&gt;&lt;/div&gt;
2956
2957
2958
2959&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2960
2961&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
296297/10
2963&lt;/p&gt;&lt;/div&gt;
2964
2965
2966
2967&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2968
2969&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2970Extracting text from Postscript
2971&lt;/p&gt;&lt;/div&gt;
2972
2973
2974
2975&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2976
2977&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2978Craig Nevill-Manning, Todd Reed, Ian H. Witten
2979&lt;/p&gt;&lt;/div&gt;
2980
2981
2982
2983&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2984
2985&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2986We show how to extract plain text from PostScript files. A textual scan is inadequate because PostScript interpreters can generate characters on the page that do not appear in the source file. Furthermore, word and line breaks are implicit in the graphical rendition, and must be inferred from the positioning of word fragments. We present a robust technique for extracting text and recognizing words and paragraphs. The method uses a standard PostScript interpreter but redefines several PostScript operators, and simple heuristics are employed to locate word and line breaks. The scheme has been used to create a full-text index, and plain-text versions, of 40,000 technical reports (34 Gbyte of PostScript). Other text-extraction systems are reviewed: none offer the same combination of robustness and simplicity.
2987&lt;/p&gt;&lt;/div&gt;
2988
2989
2990
2991&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
2992
2993&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
2994
2995&lt;/p&gt;&lt;/div&gt;
2996
2997
2998
2999&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3000
3001&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3002
3003&lt;/p&gt;&lt;/div&gt;
3004
3005
3006
3007&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3008
3009&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
301097/11
3011&lt;/p&gt;&lt;/div&gt;
3012
3013
3014
3015&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3016
3017&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3018Gathering and indexing rich fragments of the World Wide Web
3019&lt;/p&gt;&lt;/div&gt;
3020
3021
3022
3023&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3024
3025&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3026Geoffrey Holmes, William J Rogers
3027&lt;/p&gt;&lt;/div&gt;
3028
3029
3030
3031&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3032
3033&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3034While the World Wide Web (WWW) is an attractive option as a resource for teaching and research it does have some undesirable features. The cost of allowing students unlimited access can be high-both in money and time; students may become addicted to 'surfing' the web-exploring purely for entertainment-and jeopardise their studies. Students are likely to discover undesirable material because large scale search engines index sites regardless of their merit. Finally, the explosive growth of WWW usage means that servers and networks are often overloaded, to the extent that a student may gain a very negative view of the technology.
3035&lt;/p&gt;&lt;/div&gt;
3036
3037
3038
3039&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3040
3041&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3042
3043&lt;/p&gt;&lt;/div&gt;
3044
3045
3046
3047&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3048
3049&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3050We have developed a piece of software which attempts to address these issues by capturing rich fragments of the WWW onto local storage media. It is possible to put a collection onto CD ROM, providing portability and inexpensive storage. This enables the presentation of the WWW to distance learning students, who do not have internet access. The software interfaces to standard, commonly available web browsers, acting as a proxy server to the files stored on the local media, and provides a search engine giving full text searching capability within the collection.
3051&lt;/p&gt;&lt;/div&gt;
3052
3053
3054
3055&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3056
3057&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3058
3059&lt;/p&gt;&lt;/div&gt;
3060
3061
3062
3063&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3064
3065&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3066
3067&lt;/p&gt;&lt;/div&gt;
3068
3069
3070
3071&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3072
3073&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3074
3075&lt;/p&gt;&lt;/div&gt;
3076
3077
3078
3079&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3080
3081&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
308297/12
3083&lt;/p&gt;&lt;/div&gt;
3084
3085
3086
3087&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3088
3089&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3090Using model trees for classification
3091&lt;/p&gt;&lt;/div&gt;
3092
3093
3094
3095&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3096
3097&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3098Eibe Frank, Yong Wang, Stuart Inglis, Geoffrey Holmes, Ian H. Witten
3099&lt;/p&gt;&lt;/div&gt;
3100
3101
3102
3103&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3104
3105&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3106Model trees, which are a type of decision tree with linear regression functions at the leaves, form the basis of a recent successful technique for predicting continuous numeric values.  They can be applied to classification problems by employing a standard method of transforming a classification problem into a problem of function approximation.  Surprisingly, using this simple transformation the model tree inducer M5', based on Quinlan's M5, generates more accurate classifiers than the state-of-the-art decision tree learner C5.0, particularly when most of the attributes are numeric.
3107&lt;/p&gt;&lt;/div&gt;
3108
3109
3110
3111&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3112
3113&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3114
3115&lt;/p&gt;&lt;/div&gt;
3116
3117
3118
3119&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3120
3121&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3122
3123&lt;/p&gt;&lt;/div&gt;
3124
3125
3126
3127&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3128
3129&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
313097/13
3131&lt;/p&gt;&lt;/div&gt;
3132
3133
3134
3135&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3136
3137&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3138Discovering inter-attribute relationships
3139&lt;/p&gt;&lt;/div&gt;
3140
3141
3142
3143&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3144
3145&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3146Geoffrey Holmes
3147&lt;/p&gt;&lt;/div&gt;
3148
3149
3150
3151&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3152
3153&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3154It is important to discover relationships between attributes being used to predict a class attribute in supervised learning situations for two reasons.  First, any such relationship will be potentially interesting to the provider of a dataset in its own right.  Second, it would simplify a learning algorithm's search space, and the related irrelevant feature and subset selection problem, if the relationships were removed from datasets ahead of learning.  An algorithm to discover such relationships is presented in this paper.  The algorithm is described and a surprising number of inter-attribute relationships are discovered in datasets from the University of California at Irvine (UCI) repository.
3155&lt;/p&gt;&lt;/div&gt;
3156
3157
3158
3159&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3160
3161&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3162
3163&lt;/p&gt;&lt;/div&gt;
3164
3165
3166
3167&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3168
3169&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3170
3171&lt;/p&gt;&lt;/div&gt;
3172
3173
3174
3175&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3176
3177&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
317897/14
3179&lt;/p&gt;&lt;/div&gt;
3180
3181
3182
3183&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3184
3185&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3186Learning from batched data:  model combination vs data combination
3187&lt;/p&gt;&lt;/div&gt;
3188
3189
3190
3191&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3192
3193&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3194Kai Ming Ting, Boon Toh Low, Ian H. Witten
3195&lt;/p&gt;&lt;/div&gt;
3196
3197
3198
3199&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3200
3201&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3202When presented with multiple batches of data, one can either combine them into a single batch before applying a machine learning procedure or learn from each batch independently and combine the resulting models.  The former procedure, data combination, is straightforward; this paper investigates the latter, model combination.  Given an appropriate combination method, one might expect model combination to prove superior when the data in each batch was obtained under somewhat different conditions or when different learning algorithms were used on the batches.  Empirical results show that model combination often outperforms data combination even when the batches are drawn randomly from a single source of data and the same learning method is used on each.  Moreover, this is not just an artifact of one particular method of combining models: it occurs with several different combination methods.
3203&lt;/p&gt;&lt;/div&gt;
3204
3205
3206
3207&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3208
3209&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3210
3211&lt;/p&gt;&lt;/div&gt;
3212
3213
3214
3215&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3216
3217&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3218We relate this phenomenon to the learning curve of the classifiers being used.  Early in the learning process when the learning curve is steep there is much to gain from data combination, but later when it becomes shallow there is less to gain and model combination achieves a greater reduction in variance and hence a lower error rate.
3219&lt;/p&gt;&lt;/div&gt;
3220
3221
3222
3223&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3224
3225&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3226
3227&lt;/p&gt;&lt;/div&gt;
3228
3229
3230
3231&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3232
3233&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3234The practical implication of these results is that one should consider using model combination rather than data combination, especially when multiple batches of data for the same task are readily available.  It is often superior even when the batches are drawn randomly from a single sample, and we expect its advantage to increase if genuine statistical differences between the batches exist.
3235&lt;/p&gt;&lt;/div&gt;
3236
3237
3238
3239&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3240
3241&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3242
3243&lt;/p&gt;&lt;/div&gt;
3244
3245
3246
3247&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3248
3249&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3250
3251&lt;/p&gt;&lt;/div&gt;
3252
3253
3254
3255&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3256
3257&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
325897/15
3259&lt;/p&gt;&lt;/div&gt;
3260
3261
3262
3263&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3264
3265&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3266Information seeking retrieval, reading and storing behaviour of library users
3267&lt;/p&gt;&lt;/div&gt;
3268
3269
3270
3271&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3272
3273&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3274Turner K.
3275&lt;/p&gt;&lt;/div&gt;
3276
3277
3278
3279&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3280
3281&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3282In the interest of digital libraries, it is advisable that designers be aware of the potential behaviour of the users of such a system.  There are two distinct parts under investigation, the interaction between traditional libraries involving the seeking and retrieval of relevant material, and the reading and storage behaviours ensuing. Through this analysis, the findings could be incorporated into digital library facilities. There has been copious amounts of research on information seeking leading to the development of behavioural models to describe the process. Often research on the information seeking practices of individuals is based on the task and field of study. The information seeking model, presented by Ellis et al. (1993), characterises the format of this study where it is used to compare various research on the information seeking practices of groups of people (from academics to professionals). It is found that, although researchers do make use of library facilities, they tend to rely heavily on their own collections and primarily use the library as a source for previously identified information, browsing and interloan. It was found that there are significant differences in user behaviour between the groups analysed. When looking at the reading and storage of material it was hard to draw conclusions, due to the lack of substantial research and information on the topic. However, through the use of reading strategies, a general idea on how readers behave can be developed. Designers of digital libraries can benefit from the guidelines presented here to better understand their audience.
3283&lt;/p&gt;&lt;/div&gt;
3284
3285
3286
3287&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3288
3289&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3290
3291&lt;/p&gt;&lt;/div&gt;
3292
3293
3294
3295&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3296
3297&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3298
3299&lt;/p&gt;&lt;/div&gt;
3300
3301
3302
3303&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3304
3305&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
330697/16
3307&lt;/p&gt;&lt;/div&gt;
3308
3309
3310
3311&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3312
3313&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3314Proceeding of the INTERACT97 Combined Workshop on CSCW in HCI-Worldwide
3315&lt;/p&gt;&lt;/div&gt;
3316
3317
3318
3319&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3320
3321&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3322Matthias Rauterberg, Lars Oestreicher, John Grundy
3323&lt;/p&gt;&lt;/div&gt;
3324
3325
3326
3327&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3328
3329&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3330This is the proceedings for the INTERACT97 combined workshop on &amp;ldquo;CSCW in HCI-worldwide&amp;rdquo;.  The position papers in this proceedings are those selected from topics relating to HCI community development worldwide and to CSCW issues.  Originally these were to be two separate INTERACT workshops, but were combined to ensure sufficient participation for a combined workshop to run.
3331&lt;/p&gt;&lt;/div&gt;
3332
3333
3334
3335&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3336
3337&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3338
3339&lt;/p&gt;&lt;/div&gt;
3340
3341
3342
3343&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3344
3345&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3346The combined workshop has been split into two separate sessions to run in the morning of July 15&lt;sup&gt;th&lt;/sup&gt;, Sydney, Australia.  One to discuss the issues relating to the position papers focusing on general CSCW systems, the other to the development of HCI communities in a worldwide context.  The CSCW session uses as a case study a proposed groupware tool for facilitating the development of an HCI database with a worldwide geographical distribution.  The HCI community session focuses on developing the content for such a database, in order for it to foster the continued development of HCI communities.  The afternoon session of the combined workshop involves a joint discussion of the case study groupware tool, in terms of its content and likely groupware facilities.
3347&lt;/p&gt;&lt;/div&gt;
3348
3349
3350
3351&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3352
3353&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3354
3355&lt;/p&gt;&lt;/div&gt;
3356
3357
3358
3359&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3360
3361&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3362The position papers have been grouped into those focusing on HCI communities and hence content issues for a groupware database, and those focusing on CSCW and groupware issues, and hence likely groupware support in the proposed HCI database/collaboration tools.  We hope that you find the position papers in this proceedings offer a wide range of interesting reports of HCI community development worldwide, leading CSCW system research, and that a groupware tool supporting aspects of a worldwide HCI database can draw upon the varied work reported.
3363&lt;/p&gt;&lt;/div&gt;
3364
3365
3366
3367&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3368
3369&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3370
3371&lt;/p&gt;&lt;/div&gt;
3372
3373
3374
3375&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3376
3377&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3378
3379&lt;/p&gt;&lt;/div&gt;
3380
3381
3382
3383&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3384
3385&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3386
3387&lt;/p&gt;&lt;/div&gt;
3388
3389
3390
3391&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3392
3393&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
339497/17
3395&lt;/p&gt;&lt;/div&gt;
3396
3397
3398
3399&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3400
3401&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3402Internationalising a spreadsheet for Pacific Basin languages
3403&lt;/p&gt;&lt;/div&gt;
3404
3405
3406
3407&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3408
3409&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3410Robert Barbour, Alvin Yeo
3411&lt;/p&gt;&lt;/div&gt;
3412
3413
3414
3415&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3416
3417&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3418As people trade and engage in commerce, an economically dominant culture tends to migrate language into other recently contacted cultures.  Information technology (IT) can accelerate enculturation and promote the expansion of western hegemony in IT.  Equally, IT can present a culturally appropriate interface to the user that promotes the preservation of culture and language with very little additional effort.  In this paper a spreadsheet is internationalised to accept languages from the Latin-1 character set such as English, Maori and Bahasa Melayu (Malaysia's national language).  A technique that allows a non-programmer to add a new language to the spreadsheet is described.  The technique could also be used to internationalise other software at the point of design by following the steps we outline.
3419&lt;/p&gt;&lt;/div&gt;
3420
3421
3422
3423&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3424
3425&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3426
3427&lt;/p&gt;&lt;/div&gt;
3428
3429
3430
3431&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3432
3433&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3434
3435&lt;/p&gt;&lt;/div&gt;
3436
3437
3438
3439&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3440
3441&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3442
3443&lt;/p&gt;&lt;/div&gt;
3444
3445
3446
3447&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3448
3449&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
345097/18
3451&lt;/p&gt;&lt;/div&gt;
3452
3453
3454
3455&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3456
3457&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3458Localising a spreadsheet:  an Iban example
3459&lt;/p&gt;&lt;/div&gt;
3460
3461
3462
3463&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3464
3465&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3466Alvin Yeo, Robert Barbour
3467&lt;/p&gt;&lt;/div&gt;
3468
3469
3470
3471&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3472
3473&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3474Presently, there is little localisation of software to smaller cultures if it is not economically viable.  We believe software should also be localised to the languages of small cultures in order to sustain and preserve these small cultures.  As an example, we localised a spreadsheet from English to Iban.  The process in which we carried out the localisation can be used as a framework for the localisation of software to languages of small ethnic minorities.  Some problems faced during the localisation process are also discussed.
3475&lt;/p&gt;&lt;/div&gt;
3476
3477
3478
3479&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3480
3481&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3482
3483&lt;/p&gt;&lt;/div&gt;
3484
3485
3486
3487&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3488
3489&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3490
3491&lt;/p&gt;&lt;/div&gt;
3492
3493
3494
3495&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3496
3497&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3498
3499&lt;/p&gt;&lt;/div&gt;
3500
3501
3502
3503&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3504
3505&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
350697/19
3507&lt;/p&gt;&lt;/div&gt;
3508
3509
3510
3511&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3512
3513&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3514Strategies of internationalisation and localisation: a postmodernist/s perspective
3515&lt;/p&gt;&lt;/div&gt;
3516
3517
3518
3519&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3520
3521&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3522Alvin Yeo, Robert Barbour
3523&lt;/p&gt;&lt;/div&gt;
3524
3525
3526
3527&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3528
3529&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3530Many software companies today are developing software not only for local consumption but for the rest of the world.  We introduce the concepts of internationalisation and localisation and discuss some techniques using these processes.  An examination of postmodern critique with respect to the software industry is also reported.  In addition, we also feature our proposed internationalisation technique that was inspired by taking into account the researches of postmodern philosophers and mathematicians.  As illustrated in our prototype, the technique empowers non-programmers to localise their own software.  Further development of the technique and its implications on user interfaces and the future of software internationalisation and localisation are discussed.
3531&lt;/p&gt;&lt;/div&gt;
3532
3533
3534
3535&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3536
3537&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3538
3539&lt;/p&gt;&lt;/div&gt;
3540
3541
3542
3543&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3544
3545&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3546
3547&lt;/p&gt;&lt;/div&gt;
3548
3549
3550
3551&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3552
3553&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
355497/20
3555&lt;/p&gt;&lt;/div&gt;
3556
3557
3558
3559&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3560
3561&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3562Language use in software
3563&lt;/p&gt;&lt;/div&gt;
3564
3565
3566
3567&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3568
3569&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3570Alvin Yeo, Robert Barbour
3571&lt;/p&gt;&lt;/div&gt;
3572
3573
3574
3575&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3576
3577&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3578Many of the popular software we use today are in English.  Very few software applications are available in minority languages.  Besides economic goals, we justify why software should be made available to smaller cultures.  Furthermore, there is evidence that people learn and progress faster in software in their mother tongue (Griffiths et at, 1994) (Krock, 1996).  We hypothesise that experienced users of English spreadsheet can easily migrate to a spreadsheet in their native tongue i.e. Bahasa Melayu (Malaysia's national language).  Observations made in the study suggest that the native speakers of Bahasa Melayu had difficulties with the Bahasa Melayu interface.  The subjects' main difficulty was their unfamiliarity with computing terminology in Bahasa Melayu.  We present possible strategies to increase the use of Bahasa Melayu in IT.  These strategies may also be used to promote the use of other minority languages in IT.
3579&lt;/p&gt;&lt;/div&gt;
3580
3581
3582
3583&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3584
3585&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3586
3587&lt;/p&gt;&lt;/div&gt;
3588
3589
3590
3591&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3592
3593&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3594
3595&lt;/p&gt;&lt;/div&gt;
3596
3597
3598
3599&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3600
3601&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3602
3603&lt;/p&gt;&lt;/div&gt;
3604
3605
3606
3607&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3608
3609&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
361097/21
3611&lt;/p&gt;&lt;/div&gt;
3612
3613
3614
3615&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3616
3617&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3618Usability testing:  a Malaysian study
3619&lt;/p&gt;&lt;/div&gt;
3620
3621
3622
3623&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3624
3625&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3626Alvin Yeo, Robert Barbour, Mark Apperley
3627&lt;/p&gt;&lt;/div&gt;
3628
3629
3630
3631&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3632
3633&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3634An exploratory study of software assessment techniques is conducted in Malaysia.  Subjects in the study comprised staff members of a Malaysian university with a high Information Technology (IT) presence.  The subjects assessed a spreadsheet tool with a Bahasa Melayu (Malaysia's national language) interface.  Software evaluation techniques used include the think aloud method, interviews and the System Usability Scale.  The responses in the various techniques used are reported and initial results indicate idiosyncratic behaviour of Malaysian subjects.  The implications of the findings are also discussed.
3635&lt;/p&gt;&lt;/div&gt;
3636
3637
3638
3639&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3640
3641&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3642
3643&lt;/p&gt;&lt;/div&gt;
3644
3645
3646
3647&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3648
3649&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3650
3651&lt;/p&gt;&lt;/div&gt;
3652
3653
3654
3655&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3656
3657&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3658
3659&lt;/p&gt;&lt;/div&gt;
3660
3661
3662
3663&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3664
3665&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
366697/22
3667&lt;/p&gt;&lt;/div&gt;
3668
3669
3670
3671&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3672
3673&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3674Inducing cost-sensitive trees via instance-weighting
3675&lt;/p&gt;&lt;/div&gt;
3676
3677
3678
3679&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3680
3681&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3682Kai Ming Ting
3683&lt;/p&gt;&lt;/div&gt;
3684
3685
3686
3687&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3688
3689&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3690We introduce an instance-weighting method to induce cost-sensitive trees in this paper.  It is a generalization of the standard tree induction process where only the initial instance weights determine the type of tree (i.e., minimum error trees or minimum cost trees) to be induced.  We demonstrate that it can be easily adopted to an existing tree learning algorithm.
3691&lt;/p&gt;&lt;/div&gt;
3692
3693
3694
3695&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3696
3697&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3698
3699&lt;/p&gt;&lt;/div&gt;
3700
3701
3702
3703&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3704
3705&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3706Previous research gave insufficient evidence to support the fact that the greedy divide-and-conquer algorithm can effectively induce a truly cost-sensitive tree directly from the training data.  We provide this empirical evidence in this paper.  The algorithm employing the instance-weighting method is found to be comparable to or better than both C4.5 and C5 in terms of total misclassification costs, tree size and the number of high cost errors.  The instance-weighting method is also simpler and more effective in implementation than a method based on altered priors.
3707&lt;/p&gt;&lt;/div&gt;
3708
3709
3710
3711&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3712
3713&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3714
3715&lt;/p&gt;&lt;/div&gt;
3716
3717
3718
3719&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3720
3721&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3722
3723&lt;/p&gt;&lt;/div&gt;
3724
3725
3726
3727&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3728
3729&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
373097/23
3731&lt;/p&gt;&lt;/div&gt;
3732
3733
3734
3735&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3736
3737&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3738Fast convergence with a greedy tag-phrase dictionary
3739&lt;/p&gt;&lt;/div&gt;
3740
3741
3742
3743&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3744
3745&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3746Ross Peeters, Tony C. Smith
3747&lt;/p&gt;&lt;/div&gt;
3748
3749
3750
3751&lt;p&gt;&lt;div name=&quot;Body Text&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3752
3753&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3754The best general-purpose compression schemes make their gains by estimating a probability distribution over all possible next symbols given the context established by some number of previous symbols.  Such context models typically obtain good compression results for plain text by taking advantage of regularities in character sequences.  Frequent words and syllables can be incorporated into the model quickly and thereafter used for reasonably accurate prediction.  However, the precise context in which frequent patterns emerge is often extremely varied, and each new word or phrase immediately introduces new contexts which can adversely affect the compression rate
3755&lt;/p&gt;&lt;/div&gt;
3756
3757
3758
3759&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3760
3761&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3762
3763&lt;/p&gt;&lt;/div&gt;
3764
3765
3766
3767&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3768
3769&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3770A great deal of the structural regularity in a natural language is given rather more by properties of its grammar than by the orthographic transcription of its phonology.  This implies that access to a grammatical abstraction might lead to good compression.  While grammatical models have been used successfully for compressing computer programs [4], grammar-based compression of plain text has received little attention, primarily because of the difficulties associated with constructing a suitable natural language grammar.  But even without a precise formulation of the syntax of a language, there is a linguistic abstraction which is easily accessed and which demonstrates a high degree of regularity which can be exploited for compression purposes-namely, lexical categories.
3771&lt;/p&gt;&lt;/div&gt;
3772
3773
3774
3775&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3776
3777&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3778
3779&lt;/p&gt;&lt;/div&gt;
3780
3781
3782
3783&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3784
3785&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3786
3787&lt;/p&gt;&lt;/div&gt;
3788
3789
3790
3791&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3792
3793&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
379497/24
3795&lt;/p&gt;&lt;/div&gt;
3796
3797
3798
3799&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3800
3801&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3802Tag based models of English text
3803&lt;/p&gt;&lt;/div&gt;
3804
3805
3806
3807&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3808
3809&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3810W. J. Teahan, John G. Cleary
3811&lt;/p&gt;&lt;/div&gt;
3812
3813
3814
3815&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3816
3817&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3818The problem of compressing English text is important both because of the ubiquity of English as a target for compression and because of the light that compression can shed on the structure of English.  English text is examined in conjunction with additional information about the parts of speech of each word in the text (these are referred to as &amp;ldquo;tags&amp;rdquo;).  It is shown that the tags plus the text can be compressed more than the text alone.  Essentially the tags can be compressed for nothing or even a small net saving in size.  A comparison is made of a number of different ways of integrating compression of tags and text using an escape mechanism similar to PPM.  These are also compared with standard word based and character based compression programs.  The result is that the tag character and word based schemes always outperform the character based schemes.  Overall, the tag based schemes outperform the word based schemes.  We conclude by conjecturing that tags chosen for compression rather than linguistic purposes would perform even better.
3819&lt;/p&gt;&lt;/div&gt;
3820
3821
3822
3823&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3824
3825&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3826
3827&lt;/p&gt;&lt;/div&gt;
3828
3829
3830
3831&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3832
3833&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3834
3835&lt;/p&gt;&lt;/div&gt;
3836
3837
3838
3839&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3840
3841&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
384297/25
3843&lt;/p&gt;&lt;/div&gt;
3844
3845
3846
3847&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3848
3849&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3850Musical image compression
3851&lt;/p&gt;&lt;/div&gt;
3852
3853
3854
3855&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3856
3857&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3858David Bainbridge, Stuart Inglis
3859&lt;/p&gt;&lt;/div&gt;
3860
3861
3862
3863&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3864
3865&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3866Optical music recognition aims to convert the vast repositories of sheet music in the world into an on-line digital format [Bai97].  In the near future it will be possible to assimilate music into digital libraries and users will be able to perform searches based on a sung melody in addition to typical text-based searching [MSW+96].  An important requirement for such a system is the ability to reproduce the original score as accurately as possible.  Due to the huge amount of sheet music available, the efficient storage of musical images is an important topic of study.
3867&lt;/p&gt;&lt;/div&gt;
3868
3869
3870
3871&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3872
3873&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3874
3875&lt;/p&gt;&lt;/div&gt;
3876
3877
3878
3879&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3880
3881&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3882This paper investigates whether the &amp;ldquo;knowledge&amp;rdquo; extracted from the optical music recognition (OMR) process can be exploited to gain higher compression than the JBIG international standard for bi-level image compression.  We present a hybrid approach where the primitive shapes of music extracted by the optical music recognition process-note heads, note stems, staff lines and so forth-are fed into a graphical symbol based compression scheme originally designed for images containing mainly printed text.  Using this hybrid approach the average compression rate for a single page is improved by 3.5% over JBIG.  When multiple pages with similar typography are processed in sequence, the file size is decreased by 4-8%.
3883&lt;/p&gt;&lt;/div&gt;
3884
3885
3886
3887&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3888
3889&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3890
3891&lt;/p&gt;&lt;/div&gt;
3892
3893
3894
3895&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3896
3897&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3898Section 2 presents the relevant background to both optical music recognition and textual image compression.  Section 3 describes the experiments performed on 66 test images, outlining the combinations of parameters that were examined to give the best results.  The initial results and refinements are presented in Section 4, and we conclude in the last section by summarizing the findings of this work.
3899&lt;/p&gt;&lt;/div&gt;
3900
3901
3902
3903&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3904
3905&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3906
3907&lt;/p&gt;&lt;/div&gt;
3908
3909
3910
3911&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3912
3913&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3914
3915&lt;/p&gt;&lt;/div&gt;
3916
3917
3918
3919&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3920
3921&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3922
3923&lt;/p&gt;&lt;/div&gt;
3924
3925
3926
3927&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3928
3929&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
393097/26
3931&lt;/p&gt;&lt;/div&gt;
3932
3933
3934
3935&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3936
3937&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3938Correcting English text using PPM models
3939&lt;/p&gt;&lt;/div&gt;
3940
3941
3942
3943&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3944
3945&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3946W. J. Teahan, S. Inglis, J. G. Cleary, G. Holmes
3947&lt;/p&gt;&lt;/div&gt;
3948
3949
3950
3951&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3952
3953&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3954An essential component of many applications in natural language processing is a language modeler able to correct errors in the text being processed.  For optical character recognition (OCR), poor scanning quality or extraneous pixels in the image may cause one or more characters to be mis-recognized; while for spelling correction, two characters may be transposed, or a character may be inadvertently inserted or missed out.
3955&lt;/p&gt;&lt;/div&gt;
3956
3957
3958
3959&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3960
3961&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3962
3963&lt;/p&gt;&lt;/div&gt;
3964
3965
3966
3967&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3968
3969&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3970This paper describes a method for correcting English text using a PPM model.  A method that segments words in English text is introduced and is shown to be a significant improvement over previously used methods.  A similar technique is also applied as a post-processing stage after pages have been recognized by a state-of-the-art commercial OCR system.  We show that the accuracy of the OCR system can be increased from 95.9% to 96.6%, a decrease of about 10 errors per page.
3971&lt;/p&gt;&lt;/div&gt;
3972
3973
3974
3975&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3976
3977&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3978
3979&lt;/p&gt;&lt;/div&gt;
3980
3981
3982
3983&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3984
3985&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3986
3987&lt;/p&gt;&lt;/div&gt;
3988
3989
3990
3991&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
3992
3993&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
3994
3995&lt;/p&gt;&lt;/div&gt;
3996
3997
3998
3999&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4000
4001&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4002
4003&lt;/p&gt;&lt;/div&gt;
4004
4005
4006
4007&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4008
4009&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
401097/27
4011&lt;/p&gt;&lt;/div&gt;
4012
4013
4014
4015&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4016
4017&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4018Constraints on parallelism beyond 10 instructions per cycle
4019&lt;/p&gt;&lt;/div&gt;
4020
4021
4022
4023&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4024
4025&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4026John G. Cleary, Richard H. Littin, J. A. David McWha, Murray W. Pearson
4027&lt;/p&gt;&lt;/div&gt;
4028
4029
4030
4031&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4032
4033&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4034The problem of extracting Instruction Level Parallelism at levels of 10 instructions per clock and higher is considered.  Two different architectures which use speculation on memory accesses to achieve this level of performance are reviewed.  It is pointed out that while this form of speculation gives high potential parallelism it is necessary to retain execution state so that incorrect speculation can be detected and subsequently squashed.  Simulation results show that the space to store such state is a critical resource in obtaining good speedup.  To make good use of the space it is essential that state be stored efficiently and that it be retired as soon as possible.  A number of techniques for extracting the best usage from the available state storage are introduced.
4035&lt;/p&gt;&lt;/div&gt;
4036
4037
4038
4039&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4040
4041&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4042
4043&lt;/p&gt;&lt;/div&gt;
4044
4045
4046
4047&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4048
4049&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4050
4051&lt;/p&gt;&lt;/div&gt;
4052
4053
4054
4055&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4056
4057&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4058
4059&lt;/p&gt;&lt;/div&gt;
4060
4061
4062
4063&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4064
4065&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
406697/28
4067&lt;/p&gt;&lt;/div&gt;
4068
4069
4070
4071&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4072
4073&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4074Effects of re-ordered memory operations on parallelism
4075&lt;/p&gt;&lt;/div&gt;
4076
4077
4078
4079&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4080
4081&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4082Richard H. Littin, John G. Cleary
4083&lt;/p&gt;&lt;/div&gt;
4084
4085
4086
4087&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4088
4089&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4090The performance effect of permitting different memory operations to be re-ordered is examined.  The available parallelism is computed using a machine code simulator.  A range of possible restrictions on the re-ordering of memory operations is considered: from the purely sequential case where no re-ordering is permitted; to the completely permissive one where memory operations may occur in any order so that the parallelism is restricted only by data dependencies.  A general conclusion is drawn that to reliably obtain parallelism beyond 10 instructions per clock will require an ability to re-order all memory instructions.  A brief description of a feasible architecture capable of this is given.
4091&lt;/p&gt;&lt;/div&gt;
4092
4093
4094
4095&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4096
4097&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4098
4099&lt;/p&gt;&lt;/div&gt;
4100
4101
4102
4103&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4104
4105&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4106
4107&lt;/p&gt;&lt;/div&gt;
4108
4109
4110
4111&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4112
4113&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4114
4115&lt;/p&gt;&lt;/div&gt;
4116
4117
4118
4119&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;margin: 0.00mm -0.14mm 0.00mm 0.00mm;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4120
4121&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
412297/29
4123&lt;/p&gt;&lt;/div&gt;
4124
4125
4126
4127&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4128
4129&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4130OZCHI'96 Industry Session:  Sixth Australian Conference on Human-Computer Interaction
4131&lt;/p&gt;&lt;/div&gt;
4132
4133
4134
4135&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4136
4137&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4138Edited by Chris Phillips, Janis McKauge
4139&lt;/p&gt;&lt;/div&gt;
4140
4141
4142
4143&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4144
4145&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4146The idea for a specific industry session at OZCHI was first mooted at the 1995 conference in Wollongong, during questions following a session of short papers which happened (serendipitously) to be presented by people from industry.  An animated discussion took place, most of which was about how OZCHI could be made more relevant to people in industry, be it working as usability consultants, or working within organisations either as usability professionals or as `champions of the cause'.  The discussion raised more questions than answers, about the format of such as session, about the challenges of attracting industry participation, and about the best way of publishing the results.  Although no real solutions were arrived at, it was enough to place an industry session on the agenda for OZCHI'96.
4147&lt;/p&gt;&lt;/div&gt;
4148
4149
4150
4151&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4152
4153&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4154
4155&lt;/p&gt;&lt;/div&gt;
4156
4157
4158
4159&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4160
4161&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4162
4163&lt;/p&gt;&lt;/div&gt;
4164
4165
4166
4167&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4168
4169&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
417097/30
4171&lt;/p&gt;&lt;/div&gt;
4172
4173
4174
4175&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4176
4177&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4178Adaptive models of English text
4179&lt;/p&gt;&lt;/div&gt;
4180
4181
4182
4183&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4184
4185&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4186W. J. Teahan, John G. Cleary
4187&lt;/p&gt;&lt;/div&gt;
4188
4189
4190
4191&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4192
4193&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4194High quality models of English text with performance approaching that of humans is important for many applications including spelling correction, speech recognition, OCR, and encryption.  A number of different statistical models of English are compared with each other and with previous estimates from human subjects.  It is concluded that the best current models are word based with part of speech tags.  Given sufficient training text, they are able to attain performance comparable to humans.
4195&lt;/p&gt;&lt;/div&gt;
4196
4197
4198
4199&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4200
4201&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4202
4203&lt;/p&gt;&lt;/div&gt;
4204
4205
4206
4207&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4208
4209&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4210
4211&lt;/p&gt;&lt;/div&gt;
4212
4213
4214
4215&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4216
4217&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4218
4219&lt;/p&gt;&lt;/div&gt;
4220
4221
4222
4223&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4224
4225&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
422697/31
4227&lt;/p&gt;&lt;/div&gt;
4228
4229
4230
4231&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4232
4233&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4234A graphical user interface for Boolean query specification
4235&lt;/p&gt;&lt;/div&gt;
4236
4237
4238
4239&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4240
4241&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4242Steve Jones, Shona McInnes
4243&lt;/p&gt;&lt;/div&gt;
4244
4245
4246
4247&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4248
4249&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4250On-line information repositories commonly provide keyword search facilities via textual query languages based on Boolean logic.  However, there is evidence to suggest that the syntactical demands of such languages can lead to user errors and adversely affect the time that it takes users to form queries.  Users also face difficulties because of the conflict in semantics between AND and OR when used in Boolean logic and English language.  We suggest that graphical query languages, in particular Venn-like diagrams, can alleviate the problems that users experience when forming Boolean expressions with textual languages.  We describe Vquery, a Venn-diagram based user interface to the New Zealand Digital Library (NZDL).  The design of Vquery has been partly motivated by analysis of NZDL usage.  We found that few queries contain more than three terms, use of the intersection operator dominates and that query refinement is common.  A study of the utility of Venn diagrams for query specification indicates that with little or no training users can interpret and form Venn-like diagrams which accurately correspond to Boolean expressions.  The utility of Vquery is considered and directions for future work are proposed.
4251&lt;/p&gt;&lt;/div&gt;
4252
4253
4254
4255&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4256
4257&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4258
4259&lt;/p&gt;&lt;/div&gt;
4260
4261
4262
4263&lt;p&gt;&lt;div name=&quot;Normal&quot; align=&quot;left&quot; style=&quot;  padding: 0.00mm 0.00mm 0.00mm 0.00mm; &quot;&gt;
4264
4265&lt;p style=&quot;text-indent: 0.00mm; text-align: left; line-height: 4.166667mm; color: Black; background-color: White; &quot;&gt;
4266
4267&lt;/p&gt;&lt;/div&gt;
4268
4269&lt;!--Section Ends--&gt;
4270
4271
4272
4273&lt;!--
4274&lt;hr&gt;
4275&lt;address&gt;
4276&lt;a href=&quot;_httpextlink_&amp;amp;rl=0&amp;amp;href=http:%2f%2fwvware.sourceforge.net&quot;&gt;&lt;img
4277src=&quot;_httpdocimg_/wvSmall.gif&quot; height=31 width=47
4278align=left border=0 alt=&quot;wvWare&quot;&gt;&lt;/a&gt;
4279&lt;a href=&quot;_httpextlink_&amp;amp;rl=0&amp;amp;href=http:%2f%2fvalidator.w3.org%2fcheck%2freferer&quot;&gt;&lt;img
4280src=&quot;_httpdocimg_/vh40.gif&quot; height=31 width=88
4281align=right border=0 alt=&quot;Valid HTML 4.0!&quot;&gt;&lt;/a&gt;
4282Document created with &lt;a href=&quot;_httpextlink_&amp;amp;rl=0&amp;amp;href=http:%2f%2fwvware.sourceforge.net&quot;&gt;wvWare/wvWare version 1.2.4&lt;/a&gt;&lt;br&gt;
4283&lt;/address&gt;
4284--&gt;
4285
4286
4287</Content>
4288</Section>
4289</Archive>
Note: See TracBrowser for help on using the browser.