Query Expansion Strategy based on Pseudo Relevance Feedback and Term Weight Scheme for Monolingual Retrieval

Query Expansion using Pseudo Relevance Feedback is a useful technique for reformulating the query. In this paper, expansion terms are obtained by combining pseudo relevance feedback and equi-frequency partition of the documents with tf-idf scoring technique. It is observed that the groups of words that have same tf-idf score as that of query terms are better candidate words for query expansion instead of those words that are having highest tfidf. The experiment is performed on FIRE 2011 Hindi test collections, using Terrier retrieval engine. The result shows an improvement of 2.2% in mean average precision.