Table of Contents |
| List of Tables | ix |
| List of Figures | xiii |
| About the Editors | xvii |
| Contributors | xix |
| Foreword | xxi |
ch. 1 | Introduction | 1 |
1.1. | Defining the Area | 1 |
1.2. | The Content and Organization of This Book | 1 |
1.3. | The Audience for This Book | 5 |
1.4. | Further Reading | 5 |
ch. 2 | News Search Ranking | 7 |
2.1. | The Learning-to-Rank Approach | 7 |
2.1.1. | Related Works | 8 |
2.1.2. | Combine Relevance and Freshness | 8 |
2.2. | Joint Learning Approach from Clickthroughs | 10 |
2.2.1. | Joint Relevance and Freshness Learning | 12 |
2.2.2. | Temporal Features | 14 |
2.2.3. | Experiment Results | 17 |
2.2.4. | Analysis of JRFL | 19 |
2.2.5. | Ranking Performance | 24 |
2.3. | News Clustering | 27 |
2.3.1. | Architecture of the System | 29 |
2.3.2. | Offline Clustering | 30 |
2.3.3. | Incremental Clustering | 33 |
2.3.4. | Real-Time Clustering | 34 |
2.3.5. | Experiments | 37 |
| Summary | 42 |
ch. 3 | Medical Domain Search Ranking | 43 |
| Introduction | 43 |
3.1. | Search Engines for Electronic Health Records | 44 |
3.2. | Search Behavior Analysis | 47 |
3.3. | Relevance Ranking | 49 |
3.3.1. | Insights from the TREC Medical Record Track | 50 |
3.3.2. | Implementing and Evaluating Relevance Ranking in EHR Search Engines | 52 |
3.4. | Collaborative Search | 54 |
3.5. | Conclusion | 57 |
ch. 4 | Visual Search Ranking | 59 |
| Introduction | 59 |
4.1. | Generic Visual Search System | 60 |
4.2. | Text-Based Search Ranking | 61 |
4.2.1. | Text Search Models | 61 |
4.2.2. | Textual Query Preprocessing | 62 |
4.2.3. | Text Sources | 63 |
4.3. | Query Example-Based Search Ranking | 64 |
4.3.1. | Low-Level Visual Features | 64 |
4.3.2. | Distance Metrics | 65 |
4.4. | Concept-Based Search Ranking | 68 |
4.4.1. | Query-Concept Mapping | 68 |
4.4.2. | Search with Related Concepts | 70 |
4.5. | Visual Search Reranking | 71 |
4.5.1. | First Paradigm: Self-Reranking | 71 |
4.5.2. | Second Paradigm: Example-Based Reranking | 73 |
4.5.3. | Third Paradigm: Crowd Reranking | 74 |
4.5.4. | Fourth Paradigm: Interactive Reranking | 75 |
4.6. | Learning and Search Ranking | 76 |
4.6.1. | Ranking by Classification | 76 |
4.6.2. | Classification vs. Ranking | 77 |
4.6.3. | Learning to Rank | 78 |
4.7. | Conclusions and Future Challenges | 80 |
ch. 5 | Mobile Search Ranking | 81 |
| Introduction | 81 |
5.1. | Ranking Signals | 83 |
5.1.1. | Distance | 84 |
5.1.2. | Customer Reviews and Ratings | 84 |
5.1.3. | Personal Preference | 85 |
5.1.4. | Search Context: Location, Time, and Social Factors | 85 |
5.2. | Ranking Heuristics | 87 |
5.2.1. | Dataset and Experimental Setting | 88 |
5.2.2. | Customer Rating | 90 |
5.2.3. | Number of Reviews | 95 |
5.2.4. | Distance | 96 |
5.2.5. | Personal Preference | 99 |
5.2.6. | Sensitivity Analysis | 102 |
5.3. | Summary and Future Directions | 104 |
5.3.1. | Evaluation of Mobile Local Search | 104 |
5.3.2. | User Modeling and Personalized Search | 105 |
ch. 6 | Entity Ranking | 107 |
6.1. | An Overview of Entity Ranking | 107 |
6.2. | Background Knowledge | 109 |
6.2.1. | Terminology | 109 |
6.2.2. | Knowledge Base | 111 |
6.2.3. | Web Search Experience | 112 |
6.3. | Feature Space Analysis | 113 |
6.3.1. | Probabilistic Feature Framework | 113 |
6.3.2. | Graph-Based Entity Popularity Feature | 115 |
6.4. | Machine-Learned Ranking for Entities | 116 |
6.4.1. | Problem Definition | 117 |
6.4.2. | Pairwise Comparison Model | 117 |
6.4.3. | Training Ranking Function | 119 |
6.5. | Experiments | 120 |
6.5.1. | Experimental Setup | 120 |
6.5.2. | User Data-Based Evaluation | 121 |
6.5.3. | Editorial Evaluation | 124 |
6.6. | Conclusions | 125 |
ch. 7 | Multi-Aspect Relevance Ranking | 127 |
| Introduction | 127 |
7.1. | Related Work | 129 |
7.2. | Problem Formulation | 131 |
7.2.1. | Learning to Rank for Vertical Searches | 131 |
7.2.2. | Multi-Aspect Relevance Formulation | 133 |
7.2.3. | Label Aggregation | 133 |
7.2.4. | Model Aggregation | 134 |
7.3. | Learning Aggregation Functions | 135 |
7.3.1. | Learning Label Aggregation | 135 |
7.3.2. | Learning Model Aggregation | 137 |
7.4. | Experiments | 138 |
7.4.1. | Datasets | 138 |
7.4.2. | Ranking Algorithms | 140 |
7.4.3. | Offline Experimental Results | 141 |
7.4.4. | Online Experimental Results | 143 |
7.5. | Conclusions and Future Work | 145 |
ch. 8 | Aggregated Vertical Search | 147 |
| Introduction | 147 |
8.1. | Sources of Evidence | 149 |
8.1.1. | Types of Features | 149 |
8.1.2. | Query Features | 152 |
8.1.3. | Vertical Features | 153 |
8.1.4. | Vertical-Query Features | 154 |
8.1.5. | Implementation Details | 158 |
8.2. | Combination of Evidence | 158 |
8.2.1. | Vertical Selection | 158 |
8.2.2. | Vertical Presentation | 162 |
8.3. | Evaluation | 166 |
8.3.1. | Vertical Selection Evaluation | 167 |
8.3.2. | End-to-End Evaluation | 168 |
8.4. | Special Topics | 176 |
8.4.1. | Dealing with New Verticals | 176 |
8.4.2. | Explore/Exploit | 179 |
8.5. | Conclusion | 179 |
ch. 9 | Cross-Vertical Search Ranking | 181 |
| Introduction | 181 |
9.1. | The PCDF Model | 182 |
9.1.1. | Problem Formulation | 182 |
9.1.2. | Model Formulation | 183 |
9.2. | Algorithm Derivation | 186 |
9.2.1. | Objective Specification | 187 |
9.2.2. | Optimization and Implementation | 189 |
9.3. | Experimental Evaluation | 191 |
9.3.1. | Data | 192 |
9.3.2. | Experimental Setting | 193 |
9.3.3. | Results and Discussions | 193 |
9.4. | Related Work | 198 |
9.5. | Conclusions | 200 |
| References | 201 |
| Author Index | 223 |
| Subject Index | 233 |