Answering content and structure-based queries on XML documents using relevance propagation

  • Authors:

  • Karen Sauvagnat;Mohand Boughanem;Claude Chrisment

  • Affiliations:

  • IRIT/SIG, Toulouse Cedex, France;IRIT/SIG, Toulouse Cedex, France;IRIT/SIG, Toulouse Cedex, France

  • Venue:
  • Information Systems
  • Year:
  • 2006

Quantified Score

Hi-index 0.01

Visualization

Abstract

As XML documents contain both content and structure information, taking advantage of the document structure in the retrieval process can lead to better identify relevant information units. In this paper, we describe an information retrieval (IR) approach dealing with queries composed of content and structure conditions. The XFIRM model we propose is designed to be as flexible as possible to process such queries. It is based on a complete query language, derived from XPath and on a relevance values propagation method. This paper aims at evaluating functions used in the propagation process, and particularly the use of distance between nodes as a parameter. The proposed method is evaluated, thanks to the INEX evaluation initiative. Results show a relative high precision of our proposal.