On Improving Pseudo-Relevance Feedback Using Pseudo-Irrelevant Documents

Abstract. Pseudo-Relevance Feedback (PRF) assumes that the topranking n documents of the initial retrieval are relevant and extracts expansion terms from them. In this work, we introduce the notion of pseudo-irrelevant documents, i.e. high-scoring documents outside of top n that are highly unlikely to be relevant. We show how pseudo-irrelevant documents can be used to extract better expansion terms from the topranking n documents: good expansion terms are those which discriminate the top-ranking n documents from the pseudo-irrelevant documents. Our approach gives substantial improvements in retrieval performance over Model-based Feedback on several test collections. Key words: Information Retrieval, Pseudo-Relevance Feedback, Query Expansion, Pseudo-Irrelevance, Linear Classifier

• Type: Conference
• Year: 2010
• Where: ECIR
• Authors: Karthik Raman, Raghavendra Udupa, Pushpak Bhattacharyya, Abhijit Bhole
• Keywords: ECIR 2010, Expansion Terms, Information Technology, Pseudo-irrelevant Documents, Pseudo-relevance Feedback