User relevance feedback is usually utilized by Web systemsudto interpret user information needs and retrieve effectiveudresults for users. However, how to discover usefuludknowledge in user relevance feedback and how to wiselyuduse the discovered knowledge are two critical problems.udHowever, understanding what makes an individual documentudgood or bad for feedback can lead to the solution of the previous problem. In TREC 2010, we participated in the Relevance Feedback Track and experimented two models for extracting pseudo-relevance feedback to improve the ranking of retrieved documents. The first one, the main run, was a pattern-based model, whereas the second one, the optionaludrun, was a term-based model. The two models consisted of two stages: one using relevance feedback provided by TREC’10 to expand queries to extract pseudo-relevanceudfeedback; one using pseudo-relevance feedback to find usefuludpatterns and terms according to their relevance and irrelevance judgements to rank documents. In this paper, theuddetailed description of those models is presented.
展开▼
