This book constitutes the thoroughly refereed post-proceedings of the 6th International Workshop on Mining Web Data, WEBKDD 2004, held in Seattle, WA, USA in August 2004 in conjunction with the 10th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2004.
The 11 revised full papers presented together with a detailed preface went through two rounds of reviewing and improvement and were carfully selected for inclusion in the book. The extended papers are subdivided into 4 general groups: Web usage analysis and user modeling, Web personalization and recommender systems, search personalization, and semantic Web mining. The latter contains also papers from the joint KDD workshop on Mining for and from the Semantic Web, MSW 2004.
Author(s): Bamshad Mobasher, Olfa Nasraoui, Bing Liu, Brij Masand
Series: Lecture Notes in Artificial Intelligence 3932
Edition: 1
Publisher: Springer
Year: 2006
Language: English
Pages: 197
Front matter......Page 1
Introduction......Page 10
Motivation......Page 11
Related Work......Page 13
Single Node Analysis......Page 14
Sub-graph Analysis......Page 15
Whole Graph Analysis......Page 16
Our Approach......Page 17
Experiments and Results......Page 19
Single Node Level Analysis of Web Usage Graph......Page 20
Whole Graph Level Analysis of Web Usage Graph......Page 22
Conclusions......Page 24
References......Page 25
Introduction......Page 27
Related Work......Page 29
A System for Effective Web Usage Analysis......Page 30
Overall Architecture of the WUSAN System......Page 31
Preliminary Notions and Definitions......Page 33
Streams to Access Data Matrices......Page 35
A Formal Model as Basis for the LOORDSM......Page 37
Modeling the LOORDSM in UML......Page 38
A Concrete Modeling Example of a Web Warehouse......Page 40
References......Page 43
Introduction......Page 46
Architecture of a Hybrid Recommender System......Page 49
Visit Mission Identification......Page 50
Navigational Pattern Discovery......Page 52
Navigational Pattern Improved with Connectivity......Page 54
The Recommendation Engine......Page 56
Methodology......Page 57
Experimental Results......Page 58
Conclusion......Page 63
Introduction......Page 65
Context Sensitive Approach Based on Single-Step Profile Prediction Recommender System $(CSA-1-step-Rec)$......Page 67
Context Ultra-Sensitive Approach Based on Two-Step Recommender System with a Committee of Profile-Specific URL-Predictor Neural Networks $(CUSA-2-step-Rec)$......Page 70
Recommendations Based on Autoassociative Memory Hopfield Networks......Page 72
Comparative Simulation Results for $CUSA-2-step-Rec,$ CSA-1-step-Rec,$ and $K-NN$ Collaborative Filtering......Page 73
Conclusions......Page 79
References......Page 80
Introduction......Page 82
Related Work......Page 83
Factor Analysis of Dense Datasets......Page 84
Factor Analysis of Sparse Datasets......Page 85
Testing Consensus......Page 87
Testing Intrinsic Quality......Page 88
Recovering Ground Truth......Page 89
Conclusion......Page 94
Introduction......Page 96
Related Work......Page 98
Inadequacy of Joachims Algorithm......Page 99
Spy Naïve Bayes......Page 100
Optimizing Ranking Functions......Page 104
Linear Ranking Function......Page 106
Experimental Results......Page 108
Conclusions......Page 111
Introduction......Page 113
Background......Page 115
Traditional PageRank Computation......Page 116
User Profile from Internet Domains......Page 117
Design and Architecture......Page 118
Results......Page 121
Conclusions......Page 123
Introduction......Page 125
Preliminaries and Related Work......Page 127
Lexical Network-Based Features......Page 129
Proximity Features......Page 130
Likelihood Ratio Tests......Page 131
Passage Scoring and Filtering......Page 132
A Log-Linear Model......Page 133
Experiments......Page 135
Linear vs. Quadratic......Page 136
Rank Improvement Via Filtering......Page 138
Ablation and Drill-Down Studies......Page 140
Conclusion......Page 141
Introduction......Page 144
Rationale and Background......Page 145
The WebML Model and Its Supporting CASE Tool......Page 147
DEI Web Application Conceptual Logs......Page 149
MINE RULE......Page 150
Analysis of Conceptual Logs with MINE RULE......Page 151
Conclusions......Page 155
Introduction......Page 158
Preliminaries......Page 159
Conceptual Document Representation......Page 161
Boosting......Page 163
Evaluation Metrics......Page 165
Evaluation on the Reuters-21578 Corpus......Page 166
Evaluation on the OHSUMED Corpus......Page 167
Evaluation on the FAODOC Corpus......Page 171
Conclusions......Page 173
Introduction......Page 176
Related Work on Sentiments......Page 177
Problem Definition......Page 178
Bayesian Networks and Markov Blankets......Page 179
Methods: A Markov Blanket for Word Patterns......Page 180
Description of the Algorithms......Page 181
Performing the Classification......Page 184
Movie Reviews Data......Page 185
Online News Data......Page 186
Results and Analysis......Page 187
Discussion......Page 190
Future Work......Page 192
Searching for Adjacent Nodes......Page 195
Forcing a Markov Blanket DAG......Page 196
Back matter......Page 197
