Capturing multiple interests in news video retrieval by incorporating the ostensive model

Hopfgartner, F., Hannah, D., Gildea, N. and Jose, J.M. (2008) Capturing multiple interests in news video retrieval by incorporating the ostensive model. In: Second International Workshop on Personalized Access, Profile Management and Context Awareness in Databases, Auckland, New Zealand, 23 August 2008, pp. 48-55.

Abstract

We propose an adaptive news video retrieval approach which is based on the Ostensive Model of developing information needs. We therefore introduce a news video retrieval system called NewsBoy which captures the users’ implicit interactions with its graphical interface, extracts terms from visited video documents and stores them in user profiles. The terms are weighted based on the type of implicit feedback, multiple interests are identified by clustering the content of the profile. In this paper, we describe the architecture of the system and introduce our approach of adding the ostensive factor to capture the users’ evolving interest. Preliminary results show the acceptance of the system and highlights drawbacks.

• Item Type: Conference Proceedings
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Gildea, Mr Nicholas and Hannah, Mr David
• Authors: Hopfgartner, F., Hannah, D., Gildea, N., and Jose, J.M.
• Subjects: Z Bibliography. Library Science. Information Resources > Z665 Library Science. Information Science; Q Science > QA Mathematics > QA75 Electronic computers. Computer science
• College/School: College of Science and Engineering > School of Computing Science
• Research Group: Information Retrieval