Selecting effective expansion terms for diversity

Vargas, S., Santos, R.L.T., Macdonald, C. and Ounis, I. (2013) Selecting effective expansion terms for diversity. In: OAIR 2013, Lisbon, Portugal, 22-24 May 2013,

Abstract

Query expansion has been successfully applied in Information Retrieval, mostly for adhoc search tasks. On the other hand, query expansion can also fail, particularly in light of query ambiguity. For an ambiguous query, an effective strategy is to diversify the search results, in the hope of retrieving at least one relevant result for each of the possible information needs underlying the query. In this paper, we propose to tailor query expansion to diversify the search results in order to tackle query ambiguity. In particular, we introduce a novel approach to select diverse expansion terms given a suitable partition of the feedback provided by the search users. Thorough experiments in the context of the TREC 2009, 2010 and 2011 Web tracks examine the effectiveness of our approach at improving the diversification performance of state-of-the-art query expansion techniques.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Macdonald, Professor Craig and Ounis, Professor Iadh
• Authors: Vargas, S., Santos, R.L.T., Macdonald, C., and Ounis, I.
• Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
• College/School: College of Science and Engineering > School of Computing Science