Query expansion through weighted word Pairs

The aim of this paper is to propose a novel query expansion method to improve accuracy of text retrieval systems. Our technique makes use of a minimal explicit relevance feedback to expand an initial query with a structured representation called Weighted Word Pairs. Such a structure can be automatically extracted from a set of documents and uses a method for term extraction based on the probabilistic Topic Model. Evaluation has been conducted on TREC-8 repository and the proposed approach has been compared with other baseline query expansion schemes; experiments confirmed that the proposed method obtain a better performance than the baseline.