Semantic Domains in Computational Linguistics (eBook)

Semantic Domains in Computational Linguistics 2
Preface 5
Contents 7
1 Introduction 10
1.1 Lexical Semantics and Text Understanding 12
1.2 Semantic Domains: Computational Models forLexical Semantics 13
1.3 Structure of the Book 14
1.3.1 Semantic Domains 15
1.3.2 Domain Models 16
1.3.3 Semantic Domains in Text Categorization 17
1.3.4 Semantic Domains in Word Sense Disambiguation 18
1.3.5 Multilingual Domain Models 20
1.3.6 Kernel Methods for Natural Language Processing 21
2 Semantic Domains 22
2.1 The Theory of Semantic Fields 23
2.2 Semantic Fields and the meaning-is-use View 27
2.3 Semantic Domains 29
2.4 The Domain Set 31
2.5 WordNet Domains 32
2.6 Lexical Coherence: A Bridge from the Lexicon to theTexts 34
2.7 Computational Models for Semantic Domains 38
3 Domain Models 42
3.1 Domain Models: De nition 42
3.2 The Vector Space Model 43
3.3 The Domain Space 45
3.4 WordNet-Based Domain Models 47
3.5 Corpus-Based Acquisition of Domain Models 49
3.6 Latent Semantic Analysis for Term Clustering 50
3.7 The Domain Kernel 53
3.7.1 Domain Features in Supervised Learning 53
3.7.2 The Domain Kernel 55
4 Semantic Domains in Text Categorization 57
4.1 Domain Kernels for Text Categorization 57
4.1.1 Semi-supervised Learning in Text Categorization 58
4.1.2 Evaluation 59
4.1.3 Discussion 63
4.2 Intensional Learning 64
4.2.1 Intensional Learning for Text Categorization 64
4.2.2 Domain Models and the Gaussian Mixture Algorithm forIntensional Learning 66
4.2.3 Evaluation 70
4.2.4 Discussion 75
4.3 Summary 76
5 Semantic Domains in Word SenseDisambiguation 77
5.1 The Word Sense Disambiguation Task 78
5.2 The Knowledge Acquisition Bottleneck inSupervised WSD 81
5.3 Semantic Domains in the WSD Literature 82
5.4 Domain-Driven Disambiguation 84
5.4.1 Methodology 84
5.4.2 Evaluation 85
5.5 Domain Kernels for WSD 87
5.5.1 The Domain Kernel 88
5.5.2 Syntagmatic Kernels 89
5.5.3 WSD Kernels 90
5.5.4 Evaluation 90
5.6 Discussion 94
6 Multilingual Domain Models 96
6.1 Multilingual Domain Models: De nition 97
6.2 Comparable Corpora 98
6.3 The Cross-language Text Categorization Task 99
6.4 The Multilingual Vector Space Model 100
6.5 The Multilingual Domain Kernel 102
6.6 Automatic Acquisition of Multilingual DomainModels 103
6.7 Evaluation 105
6.7.1 Implementation Details 105
6.7.2 Monolingual Text Categorization Results 106
6.7.3 Cross-language Text Categorization Results 106
6.8 Summary 107
7 Conclusion and Perspectives for FutureResearch 108
7.1 Summary 108
7.2 Future Work 110
7.2.1 Consolidation of the Present Work 110
7.2.2 Domain-Driven Technologies 111
7.3 Conclusion 112
A Appendix: Kernel Methods for NaturalLanguage Processing 114
A.1 Supervised Learning 114
A.2 Feature-Based vs. Instance-Based Learning 117
A.3 Linear Classi ers 118
A.4 Kernel Methods 122
A.5 Kernel Functions 125
A.6 Kernels for Text Processing 125
References 132