Text Mining (eBook)

Preface 5
Audience 6
Supplementary Web Software 6
Acknowledgements 7
Contents 8
1 Overview of Text Mining 12
1.1 What’s Special about Text Mining? 12
1.2 What Types of Problems Can Be Solved? 17
1.3 Document Classification 18
1.4 Information Retrieval 19
1.5 Clustering and Organizing Documents 20
1.6 Information Extraction 21
1.7 Prediction and Evaluation 22
1.8 The Next Chapters 23
1.9 Historical and Bibliographical Remarks 24
2 From Textual Information to Numerical Vectors 25
2.1 Collecting Documents 25
2.2 Document Standardization 28
2.3 Tokenization 30
2.4 Lemmatization 31
2.5 Vector Generation for Prediction 35
2.6 Sentence Boundary Determination 46
2.7 Part-Of-Speech Tagging 47
2.8 Word Sense Disambiguation 49
2.9 Phrase Recognition 49
2.10 Named Entity Recognition 50
2.11 Parsing 50
2.12 Feature Generation 52
2.13 Historical and Bibliographical Remarks 54
3 Using Text for Prediction 57
3.1 Recognizing that Documents Fit a Pattern 59
3.2 How Many Documents Are Enough? 61
3.3 Document Classification 62
3.4 Learning to Predict from Text 64
3.5 Evaluation of Performance 87
3.6 Applications 91
3.7 Historical and Bibliographical Remarks 92
4 Information Retrieval and Text Mining 95
4.1 Is Information Retrieval a Form of Text Mining? 95
4.2 Key Word Search 97
4.3 Nearest-Neighbor Methods 98
4.4 Measuring Similarity 99
4.5 Web-Based Document Search 102
4.6 Document Matching 107
4.7 Inverted Lists 108
4.8 Evaluation of Performance 110
4.9 Historical and Bibliographical Remarks 111
5 Finding Structure in a Document Collection 113
5.1 Clustering Documents by Similarity 116
5.2 Similarity of Composite Documents 117
5.3 What Do a Cluster’s Labels Mean? 130
5.4 Applications 132
5.5 Evaluation of Performance 133
5.6 Historical and Bibliographical Remarks 136
6 Looking for Information in Documents 139
6.1 Goals of Information Extraction 139
6.2 Finding Patterns and Entities from Text 142
6.3 Coreference and Relationship Extraction 155
6.4 Template Filling and Database Construction 159
6.5 Applications 161
6.6 Historical and Bibliographical Remarks 164
7 Case Studies 167
7.1 Market Intelligence from the Web 167
7.2 Lightweight Document Matching for Digital Libraries 173
7.3 Generating Model Cases for Help Desk Applications 177
7.4 Assigning Topics to News Articles 182
7.5 E-mail Filtering 188
7.6 Search Engines 192
7.7 Extracting Named Entities from Documents 196
7.8 Customized Newspapers 201
7.9 Historical and Bibliographical Remarks 204
8 Emerging Directions 206
8.1 Summarization 207
8.2 Active Learning 210
8.3 Learning with Unlabeled Data 211
8.4 Different Ways of Collecting Samples 212
8.5 Question Answering 217
8.6 Historical and Bibliographical Remarks 219
Appendix: Software Notes 221
A. 1 Summary of Software 221
A.2 Requirements 222
A.3 Download Instructions 223
References 224
Author Index 236
Subject Index 240