Classification-Aware Hidden-Web Text Database Selection

1. 2.1 Database Selection Algorithms

2. 2.2 Uniform Probing for Content Summary Construction

3. 2.3 Focused Probing for Database Classification

3.1. 3. CONSTRUCTING APPROXIMATE CONTENT SUMMARIES

3.1.1. Fig 2

3.1.2. log( P ) = P 1 log( | S | ) + P 2 B = B 1 log( | S | ) + B 2

4. 3.1 Classification-Based Document Sampling

5. 3.2 Estimating Absolute Document Frequencies

5.1. 4. DATABASE SELECTION WITH SPARSE CONTENT SUMMARIES

5.1.1. Fig 6

5.1.2. Fig 7

5.1.3. Fig 8

5.1.4. Fig 9

5.1.5. Fig 10

6. 4.1 Hierarchical Database Selection

7. 4.2 Shrinkage-Based Database Selection

7.1. 5. EXPERIMENTAL SETTING

8. 5.1 Datasets

9. 5.2 Content Summary Construction Algorithms

10. 5.3 Database Selection Algorithms

10.1. 6. EXPERIMENTAL RESULTS FOR CONTENT SUMMARY QUALITY

11. 6.1 Effect of Sampling Algorithm

12. 6.2 Relationship Between Content Summaries and Categories

13. 6.3 Effect of Shrinkage

13.1. 7. EXPERIMENTAL RESULTS FOR DATABASE SELECTION ACCURACY

13.2. 8. RELATED WORK

14. 8.1 Database Selection

15. 8.2 Constructing Database Content Summaries

16. 8.3 Miscellaneous Applications of Query Probing

16.1. 9. CONCLUSION

16.2. APPENDIXES A. ESTIMATING SCORE DISTRIBUTIONS

16.3. B. ESTIMATING SCORE VARIANCE