Literature-based Discovery (eBook)

Peter Bruza, Marc Weeber (Herausgeber)

eBook Download: PDF

2008 | 2008
XII, 198 Seiten
Springer Berlin (Verlag)
978-3-540-68690-3 (ISBN)

Infos

This is the first coherent book on literature-based discovery (LBD). LBD is an inherently multi-disciplinary enterprise. The aim of this volume is to plant a flag in the ground and inspire new researchers to the LBD challenge.

Inhalt

Preface 5
Contents 6
Contributors 8
General Outlook and Possibilities 11
Literature-Based Discovery? The Very Idea 12
1 Complementarity of Hybridity and Cytology: The Birth of Cytogenetics 12
2 Suggestive Complementarity and the ABC Model 14
3 What People "Know" versus Recorded Knowledge 15
4 Fragmentation of Science 16
5 A Problem-Oriented Approach 17
6 Complementary but Disjoint Structures in the Literature of Science 18
7 Summing Up 19
7.1 Postscript: A Warning About Consequences 19
References 19
The Place of Literature-Based Discovery in Contemporary Scientific Practice 21
1 Introduction 21
2 A Case Study 22
3 Re-defining Success in LBD 25
4 Gold Standards 26
5 Concluding Remarks 28
References 28
The Tip of the Iceberg: The Quest for Innovation at the Base of the Pyramid 31
1 The Base of the Pyramid and Innovation 32
2 Disruptive Technologies 36
3 Application of Literature Based Discovery 37
4 Discovery: West to South 40
5 New Frontiers in Literature-Based Discovery 43
References 43
The ‘Open Discovery’ Challenge 46
1 Introduction 46
1.1 Text-Based Knowledge Discovery 47
1.2 General Approach 50
1.3 Previous LBD Applications 52
1.4 Improving on Co-Occurrences 53
1.5 Using History as a Guide to the Future 57
1.6 Literature Limitations 59
1.7 Integrating Gene Expression Measurements 60
2 Summary 60
References 61
Where is the Discovery in Literature-Based Discovery? 63
1 Discovery and Innovation Definitions 63
2 Radical Discovery 64
3 Literature-Based Discovery 65
4 Validating Potential Discovery (Vetting) 67
5 Examples of Validation Issues 68
5.1 Use of MeSH Variables 68
5.2 Disjointness as Sufficient Condition 72
5.3 Definition of Prior Art 75
6 Summary and Conclusions 75
References 75
Methodology and Applications 79
Analyzing LBD Methods using a General Framework 80
1 Introduction 80
2 A Framework for Analyzing LBD Research and Development 81
2.1 LBD Framework 82
3 LBD for Postulating Gene–Disease Connections 84
3.1 G2D (Perez-Iratxeta et al. 2002) 84
3.2 eVOC (Tiffin et al. 2005) 86
3.3 BITOLA (Hristovski et al. 2001, 2003) 87
3.4 Adamic et al. (2002) 88
3.5 Wilkinson and Huberman (2004) 89
3.6 Analysis of Gene-Disease LBD Approaches 90
4 General Purpose Biomedical LBD Systems 92
4.1 Manjal 94
4.2 Analysis of the General Purpose LBD Systems 95
5 Predicting Relationships from theWeb 96
5.1 Adamic and Adar (2003) 96
5.2 Ben-Dov et al. (2004) 97
5.3 Cory (1997) 97
5.4 Kumar et al. (1999) 98
5.5 Tan and Kumar (2001) 99
5.6 Bernstein et al. (2002) 99
5.7 Analysis of Web Based LBD 100
6 Conclusions 102
References 103
Evaluation of Literature-Based Discovery Systems 106
1 Introduction 106
2 Evaluation Metrics 107
2.1 Information Retrieval Metrics 107
2.2 Receiver Operating Characteristics (ROC) Curve 108
2.3 Probabilistic Approaches 109
3 Current Evaluation Approaches 109
3.1 Replicating Swanson’s Experiments 110
3.2 Using Statistical Evaluation Methods 112
3.3 Incorporating Expert Opinion 114
3.4 Publishing in the Medical Domain 115
4 User Interface Evaluation 116
5 Conclusion 117
References 117
Factor Analytic Approach to Transitive Text Mining using Medline Descriptors 119
1 Introduction 119
2 Methods 122
2.1 Data 122
2.2 Matrix Decomposition 122
2.3 Factor Screening 123
3 Results 123
3.1 Raynaud’s Disease 123
3.2 Migraine 127
3.3 Cardiac Hypertrophy 129
3.4 Multiple Sclerosis 131
4 Discussion 132
5 Conclusion 133
References 133
Literature-Based Knowledge Discovery using Natural Language Processing 136
1 Introduction 136
2 Background 137
2.1 Literature-Based Discovery 137
2.2 Natural Language Processing 139
3 Methods 142
3.1 Discovery Patterns 142
3.2 Integrated BioMedLEE and SemRep Output Format 146
4 Results 148
4.1 Rediscovering Fish Oil for Raynaud’s Disease 148
4.2 Insulin for Huntington Disease 149
4.3 Gabapentin for Parkinson’s Disease 150
4.4 Evaluation of BioMedLee and SemRep 151
5 Discussion and Further Work 152
6 Conclusions 153
References 154
Information Retrieval in Literature- Based Discovery 156
1 Introduction 156
2 Content 158
2.1 Bibliographic 158
2.2 Full-Text 159
2.3 Databases/Collections 160
3 Indexing 161
3.1 Controlled Vocabularies 162
3.2 Manual Indexing 163
3.3 Automated Indexing 165
4 Retrieval 167
4.1 Exact-Match 167
4.2 Partial-Match 168
5 Evaluation 169
5.1 System-Oriented 169
5.2 User-Oriented 171
6 IR Techniques for LBD 172
7 Conclusions 172
Suggested Readings 173
References 173
Biomedical Application of Knowledge Discovery 176
1 Overview of BioTermNet 176
2 System and Methods 178
2.1 Closed Discovery System 180
2.2 Open Discovery System 181
3 Comparison of Calculation Methods 182
4 Application to Pharmaco Genomics 187
5 Application to Linkage Analysis 188
6 Application to Microarray Analysis 191
7 Future Prospects 194
References 194
Index 196

Produkt-Details

• Erscheint lt. Verlag: 17.8.2008
• Reihe/Serie: Information Science and Knowledge Management
• Reihe/Serie: Information Science and Knowledge Management
• Zusatzinfo: XII, 198 p. 25 illus.
• Verlagsort: Berlin
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik
• Themenwelt: Technik
• Schlagworte: Data Mining • Knowledge • Knowledge Discovery • knowledge management • natural language • Natural Language Processing • Text Mining
• ISBN-10: 3-540-68690-8 / 3540686908
• ISBN-13: 978-3-540-68690-3 / 9783540686903