Muutke küpsiste eelistusi

Structure Discovery in Natural Language 2012 [Kõva köide]

  • Formaat: Hardback, 180 pages, kõrgus x laius: 235x155 mm, kaal: 471 g, XX, 180 p., 1 Hardback
  • Sari: Theory and Applications of Natural Language Processing
  • Ilmumisaeg: 09-Dec-2011
  • Kirjastus: Springer-Verlag Berlin and Heidelberg GmbH & Co. K
  • ISBN-10: 3642259227
  • ISBN-13: 9783642259227
Teised raamatud teemal:
  • Kõva köide
  • Hind: 95,02 €*
  • * hind on lõplik, st. muud allahindlused enam ei rakendu
  • Tavahind: 111,79 €
  • Säästad 15%
  • Raamatu kohalejõudmiseks kirjastusest kulub orienteeruvalt 2-4 nädalat
  • Kogus:
  • Lisa ostukorvi
  • Tasuta tarne
  • Tellimisaeg 2-4 nädalat
  • Lisa soovinimekirja
  • Formaat: Hardback, 180 pages, kõrgus x laius: 235x155 mm, kaal: 471 g, XX, 180 p., 1 Hardback
  • Sari: Theory and Applications of Natural Language Processing
  • Ilmumisaeg: 09-Dec-2011
  • Kirjastus: Springer-Verlag Berlin and Heidelberg GmbH & Co. K
  • ISBN-10: 3642259227
  • ISBN-13: 9783642259227
Teised raamatud teemal:
Current language technology is dominated by approaches that either enumerate a large set of rules, or are focused on a large amount of manually labelled data. The creation of both is time-consuming and expensive, which is commonly thought to be the reason why automated natural language understanding has still not made its way into real-life applications yet.

This book sets an ambitious goal: to shift the development of language processing systems to a much more automated setting than previous works. A new approach is defined: what if computers analysed large samples of language data on their own, identifying structural regularities that perform the necessary abstractions and generalisations in order to better understand language in the process? After defining the framework of Structure Discovery and shedding light on the nature and the graphic structure of natural language data, several procedures are described that do exactly this: let the computer discover structures without supervision in order to boost the performance of language technology applications. Here, multilingual documents are sorted by language, word classes are identified, and semantic ambiguities are discovered and resolved without using a dictionary or other explicit human input. The book concludes with an outlook on the possibilities implied by this paradigm and sets the methods in perspective to human computer interaction.

The target audience are academics on all levels (undergraduate and graduate students, lecturers and professors) working in the fields of natural language processing and computational linguistics, as well as natural language engineers who are seeking to improve their systems.
Foreword vii
Antal van den Bosch
1 Introduction
1(18)
1.1 Structure Discovery for Language Processing
1(6)
1.1.1 Structure Discovery Paradigm
3(1)
1.1.2 Approaches to Automatic Language Processing
4(1)
1.1.3 Knowledge-intensive and Knowledge-free
5(1)
1.1.4 Degrees of Supervision
6(1)
1.1.5 Contrasting Structure Discovery with Previous Approaches
7(1)
1.2 Relation to General Linguistics
7(4)
1.2.1 Linguistic Structuralism and Distributionalism
8(1)
1.2.2 Adequacy of the Structure Discovery Paradigm
9(2)
1.3 Similarity and Homogeneity in Language Data
11(2)
1.3.1 Levels in Natural Language Processing
11(1)
1.3.2 Similarity of Language Units
12(1)
1.3.3 Homogeneity of Sets of Language Units
12(1)
1.4 Vision: The Structure Discovery Machine
13(2)
1.5 Connecting Structure Discovery to NLP tasks
15(1)
1.6 Contents of this Book
16(3)
1.6.1 Theoretical Aspects of Structure Discovery
16(1)
1.6.2 Applications of Structure Discovery
17(1)
1.6.3 The Future of Structure Discovery
17(2)
2 Graph Models
19(20)
2.1 Graph Theory
19(8)
2.1.1 Notions of Graph Theory
19(4)
2.1.2 Measures on Graphs
23(4)
2.2 Random Graphs and Small World Graphs
27(12)
2.2.1 Random Graphs: Erdos-Renyi Model
27(1)
2.2.2 Small World Graphs: Watts-Strogatz Model
28(1)
2.2.3 Preferential Attachment: Barabasi-Albert Model
29(1)
2.2.4 Ageing: Power-laws with Exponential Tails
30(2)
2.2.5 Semantic Networks: Steyvers-Tenenbaum Model
32(1)
2.2.6 Changing the Power-Law's Slope: (α, β) Model
32(3)
2.2.7 Two Regimes: Dorogovtsev-Mendes Model
35(1)
2.2.8 Further Remarks on Small World Graph Models
35(2)
2.2.9 Further Reading
37(2)
3 Small Worlds of Natural Language
39(34)
3.1 Power-Laws in Rank-Frequency Distribution
39(7)
3.1.1 Word Frequency
40(1)
3.1.2 Letter N-grams
41(1)
3.1.3 Word N-grams
41(3)
3.1.4 Sentence Frequency
44(1)
3.1.5 Other Power-Laws in Language Data
44(1)
3.1.6 Modelling Language with Power-Law Awareness
45(1)
3.2 Scale-Free Small Worlds in Language Data
46(11)
3.2.1 Word Co-occurrence Graph
46(5)
3.2.2 Co-occurrence Graphs of Higher Order
51(4)
3.2.3 Sentence Similarity
55(2)
3.2.4 Summary on Scale-Free Small Worlds in Language Data
57(1)
3.3 An Emergent Random Generation Model for Language
57(16)
3.3.1 Review of Emergent Random Text Models
58(1)
3.3.2 Desiderata for Random Text Models
59(1)
3.3.3 Testing Properties of Word Streams
60(1)
3.3.4 Word Generator
60(3)
3.3.5 Sentence Generator
63(2)
3.3.6 Measuring Agreement with Natural Language
65(5)
3.3.7 Summary for the Generation Model
70(3)
4 Graph Clustering
73(28)
4.1 Review on Graph Clustering
73(10)
4.1.1 Introduction to Clustering
73(4)
4.1.2 Spectral vs. Non-spectral Graph Partitioning
77(1)
4.1.3 Graph Clustering Algorithms
77(6)
4.2 Chinese Whispers Graph Clustering
83(18)
4.2.1 Chinese Whispers Algorithm
84(4)
4.2.2 Empirical Analysis
88(3)
4.2.3 Weighting of Vertices
91(1)
4.2.4 Approximating Deterministic Outcome
92(3)
4.2.5 Disambiguation of Vertices
95(1)
4.2.6 Hierarchical Divisive Chinese Whispers
96(2)
4.2.7 Hierarchical Agglomerative Chinese Whispers
98(1)
4.2.8 Summary on Chinese Whispers
99(2)
5 Unsupervised Language Separation
101(12)
5.1 Related Work
101(1)
5.2 Method
102(1)
5.3 Evaluation
103(1)
5.4 Experiments with Equisized Parts for 10 Languages
104(3)
5.5 Experiments with Bilingual Corpora
107(2)
5.6 Case study: Language Separation for Twitter
109(2)
5.7 Summary on Language Separation
111(2)
6 Unsupervised Part-of-Speech Tagging
113(32)
6.1 Introduction to Unsupervised POS Tagging
113(1)
6.2 Related Work
114(3)
6.3 System Architecture
117(1)
6.4 Tagset 1: High and Medium Frequency Words
118(4)
6.5 Tagset 2: Medium and Low Frequency Words
122(2)
6.6 Combination of Tagsets 1 and 2
124(1)
6.7 Setting up the Tagger
125(2)
6.7.1 Lexicon Construction
125(1)
6.7.2 Constructing the Tagger
126(1)
6.7.3 Morphological Extension
127(1)
6.8 Direct Evaluation of Tagging
127(10)
6.8.1 Influence of System Components
128(3)
6.8.2 Influence of Parameters
131(1)
6.8.3 Influence of Corpus Size
132(1)
6.8.4 Domain Shifting
133(1)
6.8.5 Comparison with Clark [ 66]
134(3)
6.9 Application-based Evaluation
137(7)
6.9.1 Unsupervised POS for Supervised POS
137(2)
6.9.2 Unsupervised POS for Word Sense Disambiguation
139(2)
6.9.3 Unsupervised POS for NER and Chunking
141(3)
6.10 Conclusion on Unsupervised POS Tagging
144(1)
7 Word Sense Induction and Disambiguation
145(12)
7.1 Related Work on Word Sense Induction
145(1)
7.2 Task-oriented Definition of WSD
146(1)
7.3 Word Sense Induction using Graph Clustering
147(2)
7.3.1 Graph Clustering Parameterisation
148(1)
7.3.2 Feature Assignment in Context
149(1)
7.4 Evaluation of WSI Features in a Supervised WSD System
149(6)
7.4.1 Machine Learning Setup for Supervised WSD System
149(2)
7.4.2 SemEval-07 Lexical Sample Task
151(1)
7.4.3 Lexical Substitution System
152(1)
7.4.4 Substitution Acceptability Evaluation
153(2)
7.5 Conclusion on Word Sense Induction and Disambiguation
155(2)
8 Conclusion
157(4)
8.1 Current State of Structure Discovery
157(2)
8.2 The Future of Structure Discovery
159(2)
References 161