en
0.25
0.5
0.75
1.25
1.5
1.75
2
Rule-Based Active Sampling for Learning to Rank
Published on Nov 30, 20112784 Views
Learning to rank (L2R) algorithms rely on a labeled training set to generate a ranking model that can be later used to rank new query results. Producing these labeled training sets is usually very cos
Related categories
Chapter list
Rule-Based Active Sampling for Learning to Rank00:00
Outline00:27
Active Learning Motivation01:26
Learning to Rank using Association Rules - 104:03
Learning to Rank using Association Rules - 206:40
Relevance Estimation09:29
SSAR - Selective Sampling using Association Rules - 111:44
SSAR - Selective Sampling using Association Rules - 211:49
Experimental Setup and Results17:37
Delaying Convergence - SSARP18:42
SSARP Results20:58
Using selected instances with other L2R methods21:52
Comparing the results to LETOR baselines21:55
Comparison with other Active Learning methods22:53
Questions?23:41