An evaluation of a recursive weighing scheme for information retrieval in peer-to-peer networks

  • Authors:
  • Roberto Castiglion;Massimo Melucci

  • Affiliations:
  • University of Padua, Italy;University of Padua, Italy

  • Venue:
  • Proceedings of the 2005 ACM workshop on Information retrieval in peer-to-peer networks
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

A weighing scheme to rank ultra-peers, peers and documents is evaluated in this paper. The scheme is recursive because its components are instances of the same scheme accordingly to a hierarchy which models ultra-peers, peers and documents as informative resources. The experimental results bear evidence of the effectiveness of ultra-peer and peer ranking, i.e. visiting top-ranked ultra-peers and peers drives to relevant documents. The best results indicate that about two thirds of the relevant documents that can be retrieved by a centralized system can be obtained by exploring about one sixth of the network.