شبکه‌های اعتماد برای سیستم‌های توصیه‌گر (سیستم‌های هوش محاسباتی آتلانتیس، ۴)

This book describes research performed in the context of trust/distrust propagation and aggregation, and their use in recommender systems. This is a hot research topic with important implications for various application areas. The main innovative contributions of the work are: -new bilattice-based model for trust and distrust, allowing for ignorance and inconsistency -proposals for various propagation and aggregation operators, including the analysis of mathematical properties -Evaluation of these operators on real data, including a discussion on the data sets and their characteristics. -A novel approach for identifying controversial items in a recommender system -An analysis on the utility of including distrust in recommender systems -Various approaches for trust based recommendations (a.o. base on collaborative filtering), an in depth experimental analysis, and proposal for a hybrid approach -Analysis of various user types in recommender systems to optimize bootstrapping of cold start users. Cover......Page 1 Trust Networks for Recommender Systems......Page 4 Print: 9789491216077......Page 5 Foreword......Page 6 Preface......Page 8 Contents......Page 12 1.1 Trust Networks......Page 16 1.2 Recommender Systems......Page 19 1.3 Overview......Page 20 2 Trust Models......Page 24 2.1 Classification of Trust Models......Page 25 2.2 Trust Provenance......Page 28 2.3 Trust Score Space......Page 31 2.4 Trust Networks......Page 35 2.5 Conclusions......Page 37 3 Trust Propagation......Page 38 3.1 Setting the Scene......Page 41 3.2 Propagation Properties and Strategies......Page 43 3.2.1 Desirable Propagation Properties......Page 44 3.2.2 Propagation Patterns in Practice......Page 46 3.3.1 Preliminaries......Page 50 3.3.2 New Trust and Distrust Propagation Families......Page 53 3.4.1 The Epinions.com and CouchSurfing.org Data Sets......Page 59 3.4.2 Propagation Operators for Epinions and CouchSurfing......Page 61 3.5 Conclusions......Page 64 4.1 Aggregation Preliminaries and Context......Page 66 4.2 Trust Score Aggregation Properties......Page 70 4.3.1 Bilattice-Based Aggregation Operators......Page 76 4.3.2 Advanced Trust Score Aggregation Operators......Page 79 4.3.2.1 Fixed Weight OWA Trust Score Aggregation......Page 80 4.3.2.2 Knowledge-Enhanced KAV Trust Score Aggregation......Page 84 4.3.2.3 Knowledge-Enhanced OWA Trust Score Aggregation......Page 85 4.3.2.4 IOWA Trust Score Aggregation......Page 88 4.4 Experimental Results......Page 90 4.4.1 Comparative Analysis of the Bilattice-Based Approaches......Page 91 4.4.2 Effect of Orness Weights......Page 92 4.4.3 Impact of Knowledge Incorporation......Page 96 4.4.4 Combining Orness Weights and Knowledge Information......Page 98 4.4.5 Adaptability of the Operators......Page 100 4.4.6 Discussion......Page 102 4.5 Conclusions and Future Work......Page 104 5 Social Recommender Systems......Page 106 5.1 Classification of Recommendation Methods......Page 107 5.2 Collaborative Filtering Algorithms......Page 109 5.3 Limitations of Recommender Systems......Page 111 5.4 Evaluation of Recommender Systems......Page 112 5.4.1 Obtaining Data......Page 113 5.4.2 Evaluation Measures for Rating Prediction......Page 115 5.4.3 Evaluation Measures for Item Controversiality......Page 117 5.5 Conclusions......Page 122 6 Trust & Distrust-Based Recommendations......Page 124 6.1 Motivation......Page 125 6.2.1 State of the Art......Page 128 6.2.1.1 Mining a Trust Network......Page 129 6.2.1.2 Automatic Trust Generation......Page 137 6.2.2 Empirical Comparison......Page 140 6.2.2.1 Coverage......Page 141 6.2.2.2 Accuracy......Page 144 6.2.3 Combining Trust- and Collaborative-Based Recommendations......Page 146 6.2.4 The Impact of Propagation......Page 148 6.3.1 Distrust-Enhanced Recommendation Strategies......Page 152 6.3.2 Experimental Results......Page 157 6.4 Discussion and Future Work......Page 164 6.5 Conclusions......Page 166 7 Connection Guidancefor Cold Start Users......Page 170 7.1 The User Cold Start Problem......Page 171 7.1.1 Alleviating the Cold Start Problem......Page 172 7.1.2 Case Study for the Epinions Community......Page 173 7.2 Key Figures in a Trust-Based Recommender’s Network......Page 174 7.3 Measuring the Impact of Trusted Users......Page 177 7.3.1 Betweenness......Page 179 7.3.2 Fragmentation......Page 181 7.4 Experimental Results......Page 185 7.4.1 Contribution of Key Figures......Page 186 7.4.2 Benefit over Random Users......Page 196 7.4.3 Discussion......Page 198 7.5 Conclusions and Future Work......Page 200 8 Conclusions......Page 204 Bibliography......Page 208 Subject Index......Page 216 Annotation This book describes research performed in the context of trust/distrust propagation and aggregation, and their use in recommender systems. This is a hot research topic with important implications for various application areas. The main innovative contributions of the work are:-new bilattice-based model for trust and distrust, allowing for ignorance and inconsistency-proposals for various propagation and aggregation operators, including the analysis of mathematical properties-Evaluation of these operators on real data, including a discussion on the data sets and their characteristics.-A novel approach for identifying controversial items in a recommender system-An analysis on the utility of including distrust in recommender systems-Various approaches for trust based recommendations (a.o. base on collaborative filtering), an in depth experimental analysis, and proposal for a hybrid approach-Analysis of various user types in recommender systems to optimize bootstrapping of cold start users