Table of contents
Editors’ Forewordix
I. Invited lectures
A type-theoretic approach to anaphora and ellipsis resolution1
Human dialogue modelling using machine learning17
Learning domain theories29
Recent developments in temporal information extraction45
Annotation-based finite state processing in a large-scale NLP arhitecture61
II. Lexical semantics and lexical knowledge acquisition
Acquiring lexical paraphrases from a single corpus81
Multi-word collocation extraction by syntactic composition of collocation bigrams91
Combining independent modules in lexical multiple-choice problems101
Roget’s thesaurus and semantic similarity111
Clustering WordNet word senses121
Inducing hyperlinking rules in text collections131
Near-synonym choice in natural language generation141
III. Tagging, parsing and syntax
Fast and accurate part-of-speech tagging: The SVM approach revisited153
Part-of-speech tagging with minimal lexicalization163
Accurate annotation: An efficiency metric173
Structured parameter estimation for LFG-DOP183
Parsing without grammar — Using complete trees instead193
Phrase recognition by filtering and ranking with perceptrons205
Cascaded finite-state partial parsing: A larger-first approach217
A constraint-based bottom-up counterpart to definite clause grammars227
IV. Information extraction
Using parallel texts to improve recall in botany237
Marking atomic events in sets of related texts247
Semantically driven approach for scenario recognition in the IE system FRET257
A framework for named entity recognition in the open domain267
V. TEXT SUMMARISATION AND DOCUMENT PROCESSING
Latent semantic analysis and the construction of coherent extracts277
Facilitating email thread access by extractive summary generation287
Towards deeper understanding of the latent semantic analysis performance297
Automatic linking of similar texts across languages307
VI. OTHER NLP TOPICS
Verb phrase ellipsis detection using machine learning techniques317
HPSG-based annotation scheme for corpora development and parsing evaluation327
Arabic Morpho-syntax for Text-to-Speech337
Guessing morphological classes of unknown German nouns347
Building sense tagged corpora with volunteer contributions over the Web357
Reducing false positives by expert combination in automatic keyword indexing367
Socrates: A question answering prototype for Bulgarian377
Unsupervised natural language disambiguation using non-ambiguous words387
List of Contributors397
Index399
This article is available free of charge.