Combining Web Document Representations in a Bayesian Inference Network Model Using Link and Content-Based Evidence

  • Authors:
  • Theodora Tsikrika;Mounia Lalmas

  • Affiliations:
  • -;-

  • Venue:
  • Proceedings of the 24th BCS-IRSG European Colloquium on IR Research: Advances in Information Retrieval
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces an expressive formal Information Retrieval model developed for the Web. It is based on the Bayesian inference network model and views IR as an evidential reasoning process. It supports the explicit combination of multiple Web document representations under a single framework. Information extracted from the content of Web documents and derived from the analysis of the Web link structure is used as source of evidence in support of the ranking algorithm. This content and link-based evidential information is utilised in the generation of the multiple Web document representations used in the combination.