A cluster-based resampling method for pseudo-relevance feedback

Abstract

Typical pseudo-relevance feedback methods assume the topretrieved documents are relevant and use these pseudo-relevant documents to expand terms. The initial retrieval set can, however, contain a great deal of noise. In this paper, we present a clusterbased resampling method to select better pseudo-relevant documents based on the relevance model. The main idea is to use document clusters to find dominant documents for the initial retrieval set, and to repeatedly feed the documents to emphasize the core topics of a query. Experimental results on large-scale web TREC collections show significant improvements over the relevance model. For justification of the resampling approach, we examine relevance density of feedback documents. A higher relevance density will result in greater retrieval accuracy, ultimately approaching true relevance feedback. The resampling approach shows higher relevance density than the baseline relevance model on all collections, resulting in better retrieval accuracy in pseudo-relevance feedback. This result indicates that the proposed method is effective for pseudo-relevance feedback.

Clustering Web Search Results with Maximum Spanning Trees

“…We note that the ranking and optimality of clusters can be improved with more sophisticated techniques [12,19,20,21, inter alia]. However, this is outside the scope of this paper.…”

Mining document, concept, and term associations for effective biomedical retrieval: introducing MeSH-enhanced retrieval models

“…Under this hypothesis, a more accurate query model can be estimated from refined pseudo-feedback documents. For example, Lee et al (2008) proposed a resampling method by applying overlapping clusters to select dominant documents, which are connected with many subtopic clusters and have several highly similar documents. This resampling approach showed higher relevance density and better retrieval accuracy in their experiments.…”

Conceptual language models for domain-specific retrieval

“…The authors evaluate their algorithm on several small test collections, without achieving any improvements over the unexpanded queries. More recently, Lee, Croft, and Allan (2008) have shown that detecting clusters in a set of (pseudo-)relevant documents is helpful for identifying dominant documents for a query and, thus, for subsequent query expansion, a finding which was corroborated on different test collections by Kurland (2008). These approaches all exploit the notion that ''associations between documents convey information about the relevance of documents to requests" (Jardine & van Rijsbergen, 1971).…”