Can information retrieval systems be improved using quantum probability?

  • Authors:
  • Massimo Melucci

  • Affiliations:
  • University of Padua

  • Venue:
  • ICTIR'11 Proceedings of the Third international conference on Advances in information retrieval theory
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we reformulate the retrieval decision problem within a quantum probability framework in terms of vector subspaces rather than in terms of subsets as it is customary to state in classical probabilistic Information Retrieval. Hence we show that ranking by quantum probability of relevance in principle yields higher expected recall than ranking by classical probability at every level of expected fallout and when the parameters are estimated as accurately as possible on the basis of the available data.