Computational Linguistics and Intelligent Text Processing: Second International Conference, CICLing 2001, Mexico-City, Mexico, February 18-24, 2001. ... (Lecture Notes in Computer Science, 2004)

✍ Scribed by Alexander Gelbukh (editor)

• Publisher: Springer
• Year: 2001
• Tongue: English
• Leaves: 540
• Category: Library

✦ Synopsis

CICLing 2001 is the second annual Conference on Intelligent text processing and Computational Linguistics (hence the name CICLing), see www.CICLing.org. It is intended to provide a balanced view of the cutting edge developments in both theoretical foundations of computational linguistics and practice of natural language text processing with its numerous applications. A feature of the CICLing conferences is their wide scope that covers nearly all areas of computational linguistics and all aspects of natural language processing applications. The conference is a forum for dialogue between the specialists working in these two areas. This year our invited speakers were Graeme Hirst (U. Toronto, Canada), Sylvain Kahane (U. Paris 7, France), and Ruslan Mitkov (U. Wolverhampton, UK). They delivered excellent extended lectures and organized vivid discussions. A total of 72 submissions were received, all but very few of surprisingly high quality. After careful reviewing, the Program Committee selected for presentation 53 of them, 41 as full papers and 12 as short papers, by 98 authors from 19 countries: Spain (19 authors), Japan (15), USA (12), France, Mexico (9 each), Sweden (6), Canada, China, Germany, Italy, Malaysia, Russia, United Arab Emirates (3 each), Argentina (2), Bulgaria, The Netherlands, Ukraine, UK, and Uruguay (1 each).

✦ Table of Contents

Title Page
Copyright
Preface
Table of Contents
What Is a Natural Language and How to Describe It? Meaning-Text Approaches in Contrast with Generative Approaches
1 Introduction
2 What Is a Natural Language?
3 How to Describe a Language?
3.1 Transductive Grammars and Supercorrespondence
3.2 Example of Syntactic Transductive Grammar
3.3 Transductive Presentation in the Synthesis Direction
3.4 A Transductive Grammar in the Analysis Direction
4 Transductive Grammars and Generative Grammars
4.1 Transductive Grammars as Generative Grammars
4.2 Generative Grammars as Transductive Grammars
4.3 Equative Grammars
5 Conclusion
A Fully Lexicalized Grammar for French Based on Meaning-Text Theory
1 Introduction
2 Different Levels of Representation of a Sentence
3 Syntactic Module
3.1 Elementary Structures
3.2 Morphology-to-Syntactic Correspondence
3.3 Quasi-Dependency
4 Semantic Module
5 Extractions
5.1 Description of Extractions
5.2 Elementary Structures for Relative wh-Words
5.3 Operations of Combination with Bubbles
5.4 Comparison with Other Formalisms
6 Conclusions
Modeling the Level of Involvement of Verbal Arguments
1 Introduction: Some Reflections on Argument Structure
2 -0les as Non-primitive Concepts in Linguistic Theory
3 Calculating the Level of Involvement of Verbal Arguments
4 Discussing the Model
5 Conclusion
Magical Number Seven Plus or Minus Two: Syntactic Structure Recognition in Japanese and English Sentences
1 Introduction
2 Short-Term Memory and the 7 ± 2 Theory
3 Investigation of Japanese Sentences
4 Investigation of English Sentences
5 Conclusion
Spatio-temporal Indexing in Database Semantics
Overview
1 Intuitive Outline of Database Semantics
2 Structure-Based Storage and Retrieval
2.1 Representation of Two Concatenated Propositions
2.2 Sorting Proplets into a Network Database
2.3 Transfer of Information from the Speaker to the Hearer
3 Alternative Realizations of Propositional Content
3.1 'Railroad System' Provided by two Propositions
4 Spatio-temporal Indexing of Direct Observation
5 Reconstructing Spatio-temporal Location
5.1 Fleeting and Permanent Signs
5.2 Immediate and Mediated Reference
6 Expressing Spatio-temporal Location
7 Four Basic Types of Natural Language Communication
7.1 Basic Relations between STprop and STinter
8 Spatio-temporal Questions and Answers
8.1 Query Proplet Representing When Did Peter Cross the Street?
9 Conclusion
Russellian and Strawsonian Definite Descriptions in Situation Semantics
Introduction
1 Speakers' References and Generalized Quantifiers in Situation Semantics
1.1 Utterance Components and Speakers's References
1.2 Quantifiers in Situation Semantics
2 Russellian Account of Definites
3 Strawsonian Account of Definites in Situation Semantics
Treatment of Personal Pronouns Based on Their Parameterization
1 Introduction
2 Inflectional Categories, Co-occurrence Parameter, and Syntactic Features
3 Semantic Components
4 Formulas for Features of Plural
5 Serbo-Croatian Pronouns
6 Spanish Pronouns
7 Morpho-syntactic and Semantic Agreement
8 Translation
9 Disambiguation and Finer Specification of Spanish Pronouns
10 Conclusions
Modeling Textual Context in Linguistic Pattern Matching
Statistical Methods in Studying the Semantics of Size Adjectives
1 Introduction
2 Paradigmatic Relations between Size Adjectives
2.1 The Results of the Correlation Analysis
Numerical Model of the Strategy for Choosing Polite Expressions
1 Introduction
2 Politeness-Relationship
3 Politeness Value
3.1 The Method of Paired Comparison
3.2 Thurstone's Scaling
4 Experiments
4.1 Speech Intentions
4.2 Paired Comparison Experiment
4.3 Expression Selection Experiment
5 Statistical Analysis
5.1 Dependency on the Listener's Sex
5.2 Dependency on Speaker's Sex, Speaker's Age Bracket, and Politeness-Relationships
5.3 Interaction Between Social Distance D and Relative Power P
6 Numerical Model of Politeness-Strategy
6.1 Quantification-I Model
6.2 Linear Regression Model
7 Evaluation
7.1 The Goodness of Fitting
7.2 Prediction
8 Conclusion
Outstanding Issues in Anaphora Resolution
1 Anaphora Resolution: Where Do We Stand Now?
2 Pre-processing and Dully Automatic Anaphora Resolution
3 The Need for Annotated Corpora
4 The Resolution Algorithm Issue: Factors in Anaphora Resolution
5 Evaluation in Anaphora Resolution
6 Other Outstanding Issues
7 A Pessimistic Note: Four Traps
8 An Optimistic Voice: The Future Is Not Bleak
PHORA: A NLP System for Spanish
1 Introduction
2 The Problem
3 Resources
4 PHORA System
4.1 Method with Specification Marks to WSD
4.2 Anaphora Resolution Module
4.3 Pattern Learning Method
4.4 Anaphora Resolution Process
5 Conclusions
Belief Revision on Anaphora Resolution
A Machine-Learning Approach to Estimating the Referential Properties of Japanese Noun Phrases
1 Introduction
2 How to Estimate Referential Property
2.1 Method Used in Previous Research
2.2 The Machine-Learning Method
3 Experiment and Discussion
4 Conclusions
The Referring Expressions in the Other's Comment
Lexical Semantic Ambiguity Resolution with Bigram-Based Decision Trees
1 Introduction
2 Building a Feature Set of Bigrams
2.1 The Power Divergence Family
2.2 Dice Coefficient
3 Learning Decision Trees
4 Experimental Data
5 Experimental Method
6 Experimental Results
7 Analysis of Experimental Results
8 Discussion
9 Related Work
10 Conclusions
11 Acknowledgments
Interpretation of Compound Nominals Using WordNet
1 Introduction
2 Challenges of the Task
3 Some Recent Approaches to Noun-Noun Compound Interpretation
4 Using WordNet in Determining NN Relationships
5 Results
6 Conclusion
Specification Marks for Word Sense Disambiguation: New Development
1 Introduction
1.1 Other Approaches
2 Specification Marks Method [8, 9]
2.1 Heuristics
3 Specification Marks Method: New Development
3.1 New Heuristics
4 Empirical Results
5 Conclusion and Further Work
6 Acknowledgements
Three Mechanisms of Parser Drivingfor Structure Disambiguation
1 Introduction
2 Overview of the Model
3 Conclusions
Recent Research in the Field of Example-Based Machine Translation
Intelligent Case Based Machine Translation System
1. Introduction
2. IHSMT System
3. Lazy Learning (LL) Strategy and HMTM Learning Model
4. Algorithm of Case Based Reasoning
5. Experiment and Evaluation of the Leaning Algorithm
A Hierarchical Phrase Alignment from English and Japanese Bilingual Text
Title Generation Using a Training Corpus
1 Introduction
2 The Contrastive Title Generation Experiment
2.1 Data Description
2.2 Evaluation
2.3 Description of the Compared Approaches
2.3.1 Naive Bayesian Title Generation with Limited Vocabulary (NBL).
2.3.2 Naive Bayesian Approach with Full Vocabulary (NBF).
2.3.3 Extractive Summarization Approach Using TF/IDF (TF.IDF).
2.3.4 K Nearest Neighbor Approach (KNN).
2.4 Results and Observations
3 Conclusion and Future Work
A New Approach in Building a Corpusfor Natural Language Generation Systems
1 Introduction
2 Method
2.1 Output Text Collection
2.2 Input Determination
2.3 Text and Input Analysis
2.4 Corpus Construction
2.5 Pattern Extraction
3 Elaboration of a Corpus in the ONTOGENERATION Project
3.1 Text Collection
3.2 Input Determination
3.3 Input and Text Analysis
3.4 Corpus Construction
3.5 Pattern Extraction
3.6 Corpus Evolution
4 Conclusions and Future Work
A Study on Text Generationfrom Non-verbal Information on 2D Charts
1 Introduction
2 Analysis of a Line Chart and Accompanying Text
2.1 Analysis of the Characteristics of a Line Chart
2.2 Lexica-Grammatical Analysis
2.3 Lexical Characteristics and Line Chart Aspects
2.4 Constraints on the Usage of Words
3 The Process of Text Generation
4 Examples: A Line Chart and Generated Clauses
5 Conclusions
Interactive Multilingual Generation
1 Introduction
2 To Produce Texts in Several Languages
2.1 Machine Translation
2.2 Automatic Text Generation
2.3 Interactive Multilingual Generation
3 Interactive Generation: Technical Description
3.1 Overview
3.2 Description of the In terlingua
3.3 The First Step of the Realisation
3.4 List of Transformations with Respect to the Canonical SemR
4 A Real-World Application
4.1 Production of Weather Forecasts
4.2 Validation
4.3 Exploitation
5 Conclusion
A Computational Feature Analysis for Multilingual Character-to-Character Dialogue
1 Introduction
2 Character-to-Character Dialogue
3 Features of Character-to-Character Narrative Dialogue
3.1 The Quoted Material
3.2 The Utterer and Matrix Clause
3.3 The Surrounding Paragraphs
3.4 Direct vs. Indirect Speech
3.5 Orthography
4 NLG Aspects
5 Implementation
6 Conclusion
Experiments on Extracting Knowledge from a Machine-Readable Dictionary of Synonym Differences
1 Near-Synonyms
2 The Clustered Model of Lexical Knowledge
3 Preprocessing CTRW
4 Relating CTRW to Edmonds's Representation
4.1 Core Denotation
4.2 Denotational Distinctions
4.3 Attitudinal Distinctions
4.4 Stylistic Distinctions
5 The Class Hierarchy of Distinctions
6 The Decision-List Learning Algorithm
7 Extracting Knowledge from CTRW
7.1 Classification
7.2 Extraction
7.3 Anaphors and Comparisons
8 Results and Evaluation
9 Conclusion and Future Directions
Recognition ofAuthor's Scientific and Technical Terms
1 Introduction
2 Author's Terms in Sci-Tech Texts
2.1 Scientific Terms and Terminology
2.2 Author's Vs. Dictionary Terms
2.3 Introducing Author's Terms into Sci-Tech Texts
2.4 Distinction of Linguistic Forms
3 Automatic Text Processing and Author's Terms
4 Automatic Recognition of Author's Terms in Texts
4.1 Basic Assumptions and Key Ideas
4.2 Sketch of the Recognition Procedure
5 Conclusions and Future Work
Lexical-Semantic Tagging of an Italian Corpus
1 Introduction
2 A Brief Description of the Annotation Strategies
2.1 The ELSNET Experiment
2.2 The ISST Annotation Methodology
3 Treatment of Some Problematic Cases
3.1 Compounds
3.2 Proper Nouns
3.3 Titles
3.4 Figurative Uses and Idiomatic Expressions
4 Some Remarks about the Annotated Verbs and Argument Heads
4.1 Typical Semantic Arguments of a Verb
4.2 Verb/ Arguments Interaction at the Lexical-Semantic Level
4.3 Acquisition of Senses and Enhancement of Existing Lexical Resources
5 The Complexity of Word Sense
6 What Cannot Be Easily Encoded at the Lexical-Semantic Level of Annotation
7 Concluding Remarks
Meaning Sort -Three Examples: Dictionary Construction,Tagged Corpus Construction, and Information Presentation System -
1 Using Msort
2 Implementing Msort
3 Msort Using Different Dictionaries
3.1 Msort Using a Different 'is-a' Thesaurus
3.2 Msort Using a Dictionary Where Each Word Is Expressed witha Set of Multiple Features
4 Three Examples of Using an Msort
4.1 Dictionary Construction
4.2 Tagged Corpus Construction (Related to Semantic Similarity)
4.3 Information Retrieval
5 Conclusion
Converting Morphological Information Using Lexicalized and General Conversion
1 Introduction
2 Problems in Morphological Information Conversion
2.1 Ambiguity in Target POS
2.2 Discrepancies at Word Boundaries
3 Conversion Method
3.1 Lexicalized Conversion
4 Experiment
4.1 Lexicalized Conversion
4.2 General Conversion
4.3 Combined Conversion
5 Discussion
6 Related Work
7 Conclusion
Zipf and Heaps Laws' Coefficients Depend on Language
Applying Productive Derivational Morphologyto Term Indexing of Spanish Texts
1 Introduction
2 Morphology and Word Formation
3 Derivational Mechanisms
3.1 Non-emotive Suffixation
3.2 Back Formation
4 Phonological Conditions
5 Rules and Constraints
6 A System for Automatic Generation of Morphological Families
7 System Evaluation
8 Term Indexing Using Morphological Families
9 Conclusion
Unification-Based Lexicon and Morphology with Speculative Feature Signalling
1 Introduction
2 Open Class Words
2.1 Swedish Noun Types
2.2 The Lexicon Formalism
2.3 Paradigms for Nouns
2.4 Noun Expansion Macros
3 Inflectional Morphology
3.1 Inflectional Rules of Nouns
3.2 Choice of Formalism
4 Compositional Morphology
4.1 Feature-Enforced Branching
4.2 A Binary Switch Feature
4.3 Irregular Composition
5 Other Word Classes
5.1 Closed Class Words
5.2 Personal Pronouns
6 Evaluating Lexical Coverage
7 Conclusions
A Method of Pre-computing ConnectivityRelations for Japanese/Korean POS Tagging
1 Introduction
2 Dictionary Structure Using Trie
3 Dictionary Structure Using the AC Algorithm
4 Pre-computing Connectivity Relations
5 Evaluations
5.1 Theoretical Evaluation
5.2 Experimental Evaluation
6 Conclusion
A Hybrid Approach of Text Segmentation Based on Sensitive Word Concept for NLP
1 Introduction
2 Ambiguities in Segmentation
3 Chinese Segmentation and Sensitive Words
4 Text Segmentation System: GAOMING
4.1 Process of Text Segmentation
4.2 Analysis of Sensitive Words in a Machine-Readable Dictionary
4.3 Single-Scanning Segmentation Algorithm
5 The Detection and Resolving of Ambiguity in the Segmentation
5.1 Ambiguity Detection Algorithm
5.2 Ambiguity Resolving
6 Conclusions
7 Acknowledgments
Web-Based Arabic Morphological Analyzer
1 Introduction
2 Problem Statement
2.1 Specifications of Regular Rules
2.2 Specifications of Exception Rules
3 System Description
3.1 Rule Representation
3.1.1 Regular Rules
3.1.2 Exception Rules
3.2 Exception Lists
4 Implementation
5 Conclusion
Stochastic Parsing and Parallelism
1 Introduction
2 Extended CYK Algorithm
2.1 Non-stochastic Approach
2.2 A-Rules
2.3 Extracting Parse Trees
2.4 Dealing with Probabilities
3 Reflections on Parallelism
3.1 Distributed Memory
3.2 Shared Memory
4 Analysis of Results
5 Conclusion
Practical Nondeterministic DR(k) Parsingon Graph-Structured Stack
1 Introduction
2 Discriminating-Reverse Parsing
2.1 DR(k) Single-Action Reverse Stack Acceptors
2.2 The DR(k) Deterministic Automaton
2.3 Nondeterministic DR(k) Generation Algorithm
3 Generalized Discriminating-Reverse Parsing
3.1 Single-Decision DR(k) Parsing
3.2 GSS-Based DR{k) Parsing
3.3 Packed Shared Forest Computation
4 Conclusions
Text Categorization Using Adaptive Context Trees
1 Introduction
2 A Trade-Off in Representation
3 Probability Estimation through Adaptive Context Trees
4 Text Categorization
5 Initial Text Processing
6 Experiment on the Reuters-21578 Database
7 Experiment on the 20 N ewsgroup Database
8 Automatic Text Generation
9 Discussion
10 Conclusion
11 Annex: The Adaptive Context Tree Estimator
Text Categorization through Multistrategy Learning and Visualization
1 Introduction
2 The Individual Learning Algorithms and Their Characteristics
2.1 RIPPER Rule Learning Algorithm
2.2 Rocchio's Algorithm
2.3 RIPPER vs. Rocchio
3 A Multistrategy Approach
4 Visualization of the Output
5 Conclusions
Automatic Topic Identification Using Ontology Hierarchy
1 Introduction
2 Related Works
3 Extended Ontology
4 Automatic Topic Identification
4.1 Extraction Module
4.2 Mapping Module
4.3 Optimization Module and Main Topic
5 Experiment: Results and Analysis
6 Conclusion
Software for Creating Domain-Oriented Dictionaries and Document Clustering in Full-Text Databases
Chi-Square Classifier for Document Categorization
1 Introduction
2 Initial Data for Classification
3 The Main Algorithm
3.1 Testing of Hypotheses
3.2 Analysis of Errors of Decision
4 Conclusions and Future Work
Information Retrieval of Electronic Medical Records
1 Introduction
1.1 Description of the CPA System
2 Retrieving Relevant Information
2.1 Lexical Items
2.2 Syntactic Contexts
2.3 Semantic and Pragmatic Issues
3 Conclusion
Automatic Keyword Extraction Using Domain Knowledge
1 Introduction
2 Keyword Extraction
2.1 Manual Keyword Assignment
2.2 Automatic Indexing
2.3 Integrating the Approaches
3 Document Representation with Thesaurus Knowledge
3.1 Standard Methods: tf.idf
3.2 Thesaurus
4 Empirical Evaluation
4.1 Experimental Set-Up
4.2 Method
4.3 Experimental Results
4.5 A Second Run
5 Concluding Remarks
Approximate VLDC Pattern Matching in Shared-Forest
1 Introduction
2 The Editing Distance
3 Approximate VLDC Tree Matching
4 Approximate VLDC Matching in Shared Forest
4.1 Sharing into a Same r_keyroot
4.2 Sharing between Different r_keyroots
5 Experimental Results
Know ledge Engineering forIntelligent Information Retrieval
1 The Idea
2 The Knowledge Model
2.1 On Ontologies
2.2 On Thesauri
3 How Knowledge Is Engineered
4 Linguistic Rules
5 Application Scenarios
5.1 How SERUBA Can Improve Full-Text Retrieval
5.2 How SERUBA Can Support the Indexer
5.3 How SERUBA Can Support Automatic Indexing
6 A Final Note
Is Peritext a Key for Audiovisual Documents? The Use of Texts Describing Television Programs to Assist Indexing
1 Texts as an Access to Audiovisual Documents
2 Experiments with Different Peritexts
An Information Space Using Topic Identification for Retrieved Documents
Contextual Rules for Text Analysis
1 Introduction
2 Background
3 Contextual Rules
3.1 A Model of Text
3.2 Description of Contextual Rules
3.2.1 Syntax of Contextual Rules.
3.2.2 Meaning of Contextual Rules.
4 A Logic for Text Analysis with Contextual Rules
4.1 Previous Work
4.2 Right-Corner Chart-Parsing and Contextual Rules
4.3 A Logic for Right-Corner Inactive Chart Parsing for Contextual Rules
4.4 Management of Exclusion Sets
4.5 Search Strategies, Soundness, Completeness
5 Conclusions, Future Work
Finding Correlative Associations among News Topics*
Author Index