Evaluation of relevance and knowledge augmentation in discussion search

  • Authors:
  • Ingo Frommholz;Norbert Fuhr

  • Affiliations:
  • University of Duisburg-Essen, Duisburg, Germany;University of Duisburg-Essen, Duisburg, Germany

  • Venue:
  • ECDL'06 Proceedings of the 10th European conference on Research and Advanced Technology for Digital Libraries
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Annotation-based discussions are an important concept for today’s digital libraries and those of the future, containing additional information to and about the content managed in the digital library. To gain access to this valuable information, discussion search is concerned with retrieving relevant annotations and comments w.r.t. a given query, making it an important means to satisfy users’ information needs. Discussion search methods can make use of a variety of context information given by the structure of discussion threads. In this paper, we present and evaluate discussion search approaches which exploit quotations in different roles as highlight and context quotations, applying two different strategies, knowledge and relevance augmentation. Evaluation shows the suitability of these augmentation strategies for the task at hand; especially knowledge augmentation using both highlight and context quotations boosts retrieval effectiveness w.r.t. the given baseline.