Refereed Papers

Track: Search: Ranking and Retrieval Enhancement

Paper Title:
Ranking Refinement and Its Application to Information Retrieval

Authors:

• Rong Jin(Michigan State University)
• Hamed Valizadegan(Michigan State University)
• Hang Li(Microsoft Research Asia)

Abstract:
We consider the problem of ranking refinement, i.e., to improve the accuracy of an existing ranking function with a small set of labeled instances. We are, particularly, interested in learning a better ranking function using two complementary sources of information, ranking information given by the existing ranking function (i.e., a base ranker) and that obtained from users feedbacks. This problem is very important in information retrieval where the feedback is gradually collected. The key challenge in combining the two sources of information arises from the fact that the ranking information presented by the base ranker tends to be imperfect and the ranking information obtained from users feedbacks tends to be insufficient. We present a novel boosting framework for ranking refinement that can ffectively leverage the uses of the two sources of information. Our empirical study shows that the proposed algorithm is effective for ranking refinement, and furthermore significantly outperforms the state-of-the-arts ranking algorithms that incorporate the output from the base ranker as an additional feature of instance.

PDF version

Inquiries can be sent to: