Collaborative annotation for pseudo relevance feedback

Lioma, C., Moens, M. and Azzopardi, L. (2008) Collaborative annotation for pseudo relevance feedback. In: ECIR’08 Workshop on: Exploiting Semantic Annotations for Information Retrieval, Glasgow, Scotland, 30 Mar 2008,

Abstract

We present a pseudo relevance feedback technique for infor- mation retrieval, which expands keyword queries with semantic annota- tion found in the freely available Del.icio.us collaborative tagging system. We hypothesise that collaborative tags represent semantic information that may render queries more informative, and hence enhance retrieval performance. Experiments with three different techniques of enriching queries with Del.icio.us tags, and also varying the number of tags used for expansion between 1-10, show small improvement in retrieval preci- sion, over a baseline of short keyword queries.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Azzopardi, Dr Leif
• Authors: Lioma, C., Moens, M., and Azzopardi, L.
• Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
• College/School: College of Science and Engineering > School of Computing Science
• Journal Name: feedback