Herrera-Viedma / Pasi / Crestani Soft Computing in Web Information Retrieval

Models and Applications
1. Auflage 2008
ISBN: 978-3-540-31590-2
Verlag: Springer Berlin Heidelberg
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, Band 197, 319 Seiten

Reihe: Studies in Fuzziness and Soft Computing

Appl.Mathematics/Computational Methods of Engineering - Short description currently not available.

Autoren/Hrsg.

Herrera-Viedma, Enrique

Pasi, Gabriella

Crestani, Fabio

Inhaltsverzeichnis

1;Preface;6
2;Acknowledgments;9
3;Foreword;10
4;Contents;11
5;Part I Document Classification;13
5.1;A Dynamic Hierarchical Fuzzy Clustering Algorithm for Information Filtering;14
5.1.1;1 Introduction;15
5.1.2;2 Categorization of Documents in IR Based;16
5.1.3;on Clustering Techniques;16
5.1.3.1;2.1 Partitioning Clustering Method;17
5.1.3.2;2.2 Hierarchical Clustering Method;20
5.1.3.3;divisive;20
5.1.3.4;agglomerative;20
5.1.3.5;2.3 Incremental Clustering Method;20
5.1.3.6;single-Pass;21
5.1.3.7;K-nearest neighbour;21
5.1.4;3 The Rationale of the Proposed Approach;21
5.1.4.1;fuzzy partition;22
5.1.4.2;fuzzy hierarchy;22
5.1.4.3;incremental dynamic clustering;22
5.1.4.4;algorithm).;22
5.1.4.5;semi-supervised fuzzy clustering;23
5.1.4.6;techniques;23
5.1.4.7;documents are represented by large and sparse vectors;23
5.1.4.8;number of clusters that must be generated at each;23
5.1.4.9;level of the hierarchy is not known;23
5.1.5;4 The Dynamic Fuzzy Hierarchical Clustering Algorithm;24
5.1.5.1;4.1 Documents Indexing Criteria for Clustering Purposes;24
5.1.5.2;Den;25
5.1.5.3;4.2 Input of the Clustering Algorithm;25
5.1.5.4;4.3 The Fuzzy Hierarchy of Documents;26
5.1.5.5;4.4 Generation of the Fuzzy Clusters;27
5.1.5.6;Histogram;28
5.1.5.7;Histogram(;29
5.1.5.8;4.5 Updating the Fuzzy Hierarchy with New Documents;29
5.1.6;5 Preliminary Experiments;30
5.1.6.1;sport;30
5.1.6.2;safety;30
5.1.6.3;game;30
5.1.6.4;math;30
5.1.6.5;lego;30
5.1.6.6;math;30
5.1.6.7;game ,;30
5.1.6.8;sport;30
5.1.6.9;game );30
5.1.6.10;Histogram;31
5.1.6.11;lego;31
5.1.7;6 Conclusions;32
5.1.8;Acknowledgements;32
5.1.9;References;32
5.2;A Theoretical Framework for Web Categorization in Hierarchical Directories using Bayesian Networks;35
5.2.1;1 Introduction;35
5.2.2;2 Introduction to Bayesian Networks;37
5.2.3;3 Related Work on Hierarchical Categorization;38
5.2.4;4 Representing Hierarchical Web Directories using a Bayesian Network;39
5.2.4.1;4.1 Improving the Basic Model;43
5.2.4.2;4.2 Assessment of the Probability Distributions;44
5.2.5;5 Categorizing Web Pages: Inference;47
5.2.5.1;ci,;48
5.2.5.2;cj);48
5.2.5.3;ci, cj;48
5.2.5.4;ci;48
5.2.5.5;cj;48
5.2.5.6;ci,;48
5.2.5.7;cj);48
5.2.5.8;ci;48
5.2.5.9;cj;48
5.2.5.10;Example:;49
5.2.6;6 Concluding Remarks and Future Works;51
5.2.7;Acknowledgments;52
5.2.8;References;52
5.3;Personalized Knowledge Models Using RDF-Based Fuzzy Classi.cation;54
5.3.1;1 Introduction;54
5.3.2;2 Preliminary Considerations: From a Human to Machine-Oriented Vision of Information;56
5.3.2.1;2.1 Usage Scenario;58
5.3.3;3 Architecture Overview;58
5.3.3.1;3.1 Work.ow Scenario;61
5.3.4;4 Knowledge Acquisition;61
5.3.5;5 Features Extraction;62
5.3.5.1;5.1 Relevance Measurement of the Features;62
5.3.5.2;Collection of RDF pages:;63
5.3.5.3;Collection of schemas or dictionaries:;63
5.3.5.4;Dictionaries related to the current RDF page:;63
5.3.5.5;Accuracy:;63
5.3.5.6;Instance Relevance:;64
5.3.5.7;Property Relevance:;64
5.3.6;6 Rule-based Classi.cation;66
5.3.6.1;6.1 Clustering of RDF Pages;66
5.3.6.2;6.2 Rules Generation;67
5.3.6.3;If;67
5.3.6.4;then;67
5.3.7;7 Experimental Results;69
5.3.8;8 Conclusions;71
5.3.9;References;72
5.4;A Genetic Programming Approach for Combining Structural and Citation-Based Evidence for Text Classi.cation in Web Digital Libraries;74
5.4.1;1 Introduction;74
5.4.2;2 Background;76
5.4.3;3 Our Approach;77
5.4.3.1;3.1 GP System Con.gurations;77
5.4.3.2;Algorithm 1:;77
5.4.3.3;3.2 Used Terminals;78
5.4.3.4;Structural Similarity Measures;78
5.4.3.5;Citation-based Similarity Measures;79
5.4.4;4 The Framework for Classi.cation;81
5.4.5;5 Experiments;82
5.4.5.1;5.1 Sampling;83
5.4.5.2;5.2 Baselines;84
5.4.5.3;5.3 Experimental Set Up;85
5.4.5.4;5.4 Experimental Results;85
5.4.6;6 Related Work;87
5.4.7;7 Conclusion;88
5.4.8;Acknowledgements;88
5.4.9;References;89
6;Part II Semantic Web;93
6.1;Adding a Trust Layer to Semantic Web Metadata;94
6.1.1;1 Introduction;94
6.1.2;2 The Architecture;95
6.1.2.1;2.1 The Metadata Format;97
6.1.2.2;2.2 Modelling User Behavior in Implicit Voting;100
6.1.2.3;2.3 Trust Assertions’ Format;101
6.1.3;3 The Reputation Computation Problem;103
6.1.3.1;3.1 Choice of the Aggregation Operator;104
6.1.3.2;De.nition 1.;104
6.1.3.3;A1;105
6.1.3.4;A2;105
6.1.3.5;A3;105
6.1.3.6;p;105
6.1.3.7;w;105
6.1.3.8;3.2 The WOWA Operator;105
6.1.3.9;De.nition 2.;105
6.1.3.10;p;105
6.1.3.11;w;105
6.1.3.12;p;105
6.1.3.13;w;105
6.1.3.14;p;106
6.1.3.15;w,;106
6.1.3.16;a,;106
6.1.3.17;w;106
6.1.3.18;a;106
6.1.3.19;w;106
6.1.3.20;p;106
6.1.3.21;w;106
6.1.3.22;p;106
6.1.3.23;w;107
6.1.3.24;p;107
6.1.3.25;3.3 An Example;107
6.1.3.26;w;107
6.1.3.27;The Di.dent Approach;107
6.1.3.28;a;107
6.1.3.29;w;107
6.1.3.30;wn;107
6.1.3.31;p=[;108
6.1.3.32;The Con.dent Approach;109
6.1.3.33;a;109
6.1.3.34;wn;109
6.1.4;4 Conclusions;109
6.1.5;Acknowledgments;110
6.1.6;References;110
6.2;A Fuzzy Linguistic Multi-agent Model Based on Semantic Web Technologies and User Pro.les;112
6.2.1;1 Introduction;112
6.2.2;2 Methodological Tools;114
6.2.2.1;2.1 Fuzzy Linguistic Tools;114
6.2.2.2;2.2 Filtering Techniques;115
6.2.2.3;2.3 Semantic Web Technologies;116
6.2.3;3 The Fuzzy Linguistic Multi-agent Model Based on Semantic Web and User Pro.les;116
6.2.3.1;Semantic Retrieval Phase:;117
6.2.3.2;Feedback Phase:;118
6.2.3.3;3.1 Feedback Phase: User Pro.le Updating Process;118
6.2.3.4;Step 1:;121
6.2.3.5;Step 2:;121
6.2.3.6;Step 3:;121
6.2.3.7;3.2 Feedback Phase: Recommendation Process;121
6.2.3.8;Step 1:;123
6.2.3.9;Step 2:;123
6.2.4;4 Example;123
6.2.5;5 Concluding Remarks;124
6.2.6;References;125
6.3;Fuzzy Concept-Based Models in Information Browsers;128
6.3.1;1 Introduction;128
6.3.2;2 Fuzziness in Concept Browsers;130
6.3.2.1;2.1 A Basic Model for Resources and Their Annotations;130
6.3.2.2;2.2 Fuzzy Elements in the Implementation;134
6.3.2.3;of Ontology-Guided Tactics;134
6.3.3;3 A Semantic Web Implementation of a Fuzzy Generic Concept Browser;134
6.3.4;4 Conclusions and Future Work;139
6.3.5;References;139
6.4;Evaluation of Term-based Queries using Possibilistic Ontologies;142
6.4.1;1 Introduction;142
6.4.2;2 From Fuzzy to Qualitative Pattern Matching;143
6.4.2.1;2.1 Fuzzy Pattern Matching;143
6.4.2.2;2.2 Possibilistic Ontology;145
6.4.2.3;2.3 Qualitative Pattern Matching;148
6.4.2.4;2.4 Other Approaches Using Ontologies;149
6.4.3;3 Using Qualitative Pattern Matching on a Database;151
6.4.3.1;Description of the;151
6.4.3.2;Platform;151
6.4.3.3;Used Ontologies;151
6.4.3.4;Examples of Queries;153
6.4.4;4 Retrieving Titles Using Qualitative Pattern Matching;157
6.4.4.1;4.1 Data Description;158
6.4.4.2;4.2 Examples of Queries;158
6.4.4.3;4.3 Evaluation and Results;159
6.4.5;5 Toward an Extension of the Approach to Full-text IR;163
6.4.5.1;5.1 Possibilistic Indexing;163
6.4.5.2;5.2 Query Evaluation;164
6.4.6;6 Conclusion;165
6.4.7;References;166
7;Part III Web Information Retrieval;168
7.1;Formal Theory of Connectionist Web Retrieval;169
7.1.1;1 Introduction;169
7.1.2;2 Connectionist Web Information Retrieval;171
7.1.2.1;2.1 Arti.cial Neural Network;171
7.1.2.2;Theorem 1.;172
7.1.2.3;2.2 Information Retrieval Using Multi-Layered;174
7.1.2.4;Arti.cial Neural Networks;174
7.1.2.5;2.3 Arti.cial Neural Network-based Web Retrieval –;177
7.1.2.6;A Literature Overview;177
7.1.3;3 Formal Theory of Connectionist Web Retrieval;180
7.1.3.1;3.1 PageRank;180
7.1.3.2;3.2 Authorities and Hubs;181
7.1.3.3;3.3 Interaction Information Retrieval;182
7.1.3.4;3.4 Interaction Information Retrieval: Particular Case;183
7.1.3.5;of the Generic Equation;183
7.1.3.6;3.5 PageRank: Particular Case of the Generic Equation;185
7.1.3.7;3.6 Hubs and Authorities: Particular Case;186
7.1.3.8;of the Generic Equation;186
7.1.4;4 Computational Complexity;187
7.1.5;in Connectionist Web Retrieval;187
7.1.5.1;4.1 Basic Concepts;187
7.1.5.2;4.2 Computational Complexity in Soft Computing-based;188
7.1.5.3;Information Retrieval – A Literature Overview;188
7.1.5.4;4.3 Computational Complexity;192
7.1.5.5;of Winner-Take-All-based Retrieval;192
7.1.5.6;Theorem 2.;192
7.1.5.7;Theorem 3.;192
7.1.5.8;Theorem 4.;192
7.1.5.9;Theorem 5.;193
7.1.6;5 Conclusions;195
7.1.7;Acknowledgements;196
7.1.8;References;196
7.2;Semi-fuzzy Quantifiers for Information Retrieval;201
7.2.1;1 Introduction;201
7.2.2;2 Related Work;203
7.2.3;3 Semi-fuzzy Quanti.ers for Information Retrieval;204
7.2.3.1;De.nition 1 (fuzzy quanti.er).;204
7.2.3.2;De.nition 2 (semi-fuzzy quanti.er).;205
7.2.3.3;De.nition 3 (quanti.er fuzzi.cation mechanism).;206
7.2.3.4;De.nition 4 (;206
7.2.3.5;cut).;206
7.2.3.6;3.1 Query Language;208
7.2.3.7;3.2 Semantics;208
7.2.3.8;3.3 Example;209
7.2.4;4 Semi-fuzzy Quanti.ers and OWA Quanti.cation;210
7.2.4.1;4.1 Linguistic Quanti.cation using OWA Operators;210
7.2.4.2;4.2 Linguistic Quanti.cation using SFQ;211
7.2.4.3;4.3 Remarks;213
7.2.5;5 Experiments;214
7.2.5.1;5.1 Experiments:;216
7.2.5.2;5.2 Experiments: Pivoted Document Length Normalization;220
7.2.6;6 Conclusions and Further Work;222
7.2.7;Acknowledgements;223
7.2.8;References;223
7.2.9;Appendix A;225
7.2.9.1;weights;226
7.2.9.2;pivoted weights;226
7.3;Helping Users in Web Information Retrieval Via Fuzzy Association Rules;227
7.3.1;1 Introduction;227
7.3.2;2 Query Re.nement;228
7.3.3;3 Association Rules and Fuzzy Association Rules;229
7.3.3.1;3.1 Association Rules;230
7.3.3.2;3.2 Fuzzy Association Rules;230
7.3.3.3;3.3 Measures for Association and Fuzzy Association Rules;231
7.3.4;4 Query Re.nement via Fuzzy Association Rules;232
7.3.5;5 Document Representation for Association Rule Extraction;233
7.3.5.1;5.1 Text Transactions;234
7.3.5.2;Fuzzy Text Transactions;234
7.3.6;6 Extraction of Fuzzy Association Rules;235
7.3.6.1;Algorithm 1;235
7.3.6.2;6.1 The Selection of Terms for Query Re.nement;236
7.3.7;7 Experimental Examples;237
7.3.8;8 Conclusions and Future Work;240
7.3.9;Acknowledgements;240
7.3.10;References;240
7.4;Combining Soft and Hard Techniques for the Analysis of Batch Retrieval Tasks;244
7.4.1;1 Introduction: a Prototypical Batch Retrieval Task;244
7.4.1.1;The Statement of the Task;245
7.4.1.2;Making Some Hypotheses Explicit;246
7.4.1.3;Hypothesis 1 (Conjunctive querying).;246
7.4.1.4;Hypothesis 2 (Disjunctive querying).;247
7.4.1.5;A Decomposition of the Problem;247
7.4.1.6;Problem 1.;247
7.4.1.7;Problem 2.;248
7.4.1.8;Problem 3.;248
7.4.2;2 A Perfect Information Retrieval System;248
7.4.2.1;Galois Connections and Adjunctions;248
7.4.2.2;De.nition 1.;248
7.4.2.3;Proposition 1 (Polarity).;248
7.4.2.4;Proposition 2 (Axiality).;249
7.4.2.5;The Polarity of Conjunctive Querying;249
7.4.2.6;Theorem 1 (Basic theorem on Concept Lattices ([8], p. 20)).;250
7.4.2.7;Gathering Results: the Solution to Problem 1;250
7.4.3;3 Relevance-Induced Analysis of Retrieval Systems;251
7.4.3.1;Equivalences and Partitions De.ned by Relevance Relations;252
7.4.3.2;Rough Set Analysis of Relevance;252
7.4.3.3;An Example: the Relevance Lattice of a TREC Task;254
7.4.4;4 Designing the Description Mappings;256
7.4.4.1;Constraints for the Description Mappings;256
7.4.4.2;A Solution: Infomorphisms;257
7.4.5;5 Related Work and Discussion;258
7.4.6;Acknowledgements;260
7.4.7;References;260
8;Part IV Web Application;262
8.1;Search Advertising;263
8.1.1;1 Introduction;263
8.1.2;2 Basic Concepts;266
8.1.2.1;2.1 Keyword-targeted Advertising;266
8.1.2.2;2.2 Content-targeted Advertising;267
8.1.2.3;2.3 The Search Advertising Network;268
8.1.2.4;The Users;270
8.1.2.5;The Advertisers;271
8.1.2.6;The Publishers;271
8.1.3;3 Search Advertising Systems;272
8.1.3.1;3.1 Relevance Matching;274
8.1.3.2;dj;277
8.1.3.3;dj);277
8.1.3.4;dj).;278
8.1.3.5;3.2 Ranking;279
8.1.3.6;3.3 Fraud Detection;281
8.1.3.7;3.4 Measurements and Feedback;282
8.1.4;4 Conclusions;285
8.1.5;References;285
8.2;Information Loss in Continuous Hybrid Microdata: Subdomain-Level Probabilistic Measures;290
8.2.1;1 Introduction;290
8.2.1.1;1.1 Contribution and Plan of This Paper;291
8.2.2;2 A Low-cost Method for Hybrid Microdata Generation;292
8.2.2.1;Algorithm 1 (Basic Procedure);292
8.2.2.2;Algorithm 2 (Modi.cation of Matrix;293
8.2.3;3 Properties of the Proposed Scheme;294
8.2.3.1;3.1 Performance and Complexity;294
8.2.3.2;3.2 Data Utility;294
8.2.4;4 A Generic Information Loss Measure;295
8.2.5;5 Empirical Work;296
8.2.5.1;5.1 Information Loss and Disclosure Risk Measures;296
8.2.5.2;5.2 The Data Set;296
8.2.5.3;5.3 The Results;297
8.2.5.4;Results on the Overall Dataset;297
8.2.5.5;Top-down Generation: Posterior Subdomains;297
8.2.5.6;Bottom-up Generation: Prior Subdomains;298
8.2.6;6 Conclusions and Future Research;298
8.2.7;Acknowledgments;300
8.2.8;References;300
8.3;Access to a Large Dictionary of SpanishSynonyms: A Tool for Fuzzy InformationRetrieval.;302
8.3.1;1 Introduction;302
8.3.2;2 A Short Historical Introduction to Synonymy;306
8.3.3;3 A Computational View of Synonymy;307
8.3.4;4 General Architecture of an Electronic Dictionary;310
8.3.5;of Synonyms;310
8.3.6;5 Improving the Dictionary;312
8.3.7;6 Stand-alone Use of FDSA;314
8.3.7.1;The electronic dictionaries.;315
8.3.7.2;The algorithms that calculate the degrees of synonymy and;315
8.3.7.3;antonymy.;315
8.3.7.4;The graphical user interface.;316
8.3.8;7 Conclusions;317
8.3.9;References;318
9;More eBooks at www.ciando.com;0