Measuring the informativeness of a retrieval process

  • Authors:
  • Jean Tague-Sutcliffe

  • Affiliations:
  • School of Library and Information Science, University of Western Ontario, London, Ontario, Canada N6G 1H1

  • Venue:
  • SIGIR '92 Proceedings of the 15th annual international ACM SIGIR conference on Research and development in information retrieval
  • Year:
  • 1992

Quantified Score

Hi-index 0.00

Visualization

Abstract

Evaluation of information retrieval systems should be based on measures of the information provided by the retrieval process, “informativeness” measures which take into account the interactive and full-text nature of present-day systems and the different types of questions which are asked of them. Desirable properties for an informativeness measure are developed, including context sensitivity, user centrality, and logarithmic response. A hypergraph-based framework for measuring the informativeness of a retrieval process is presented and a measure developed which satisfies the desired properties. The measure is compared to previously developed information measures and illustrated via an application.