Peer-to-Peer Keyword Search Using Keyword Relationship

  • Authors:
  • Kiyohide Nakauchi;Yuichi Ishikawa;Hiroyuki Morikawa;Tomonori Aoyama

  • Affiliations:
  • -;-;-;-

  • Venue:
  • CCGRID '03 Proceedings of the 3st International Symposium on Cluster Computing and the Grid
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Decentralized and unstructured peer-to-peer (P2P) networkssuch as Gnutella are attractive for Internet-scale informationretrieval and search systems because they requireneither any centralized directory nor any centralized managementof overlay network topology and data placement.However, due to this decentralized architecture, currentP2P keyword search systems lack useful global knowledgesuch as popularity of data items and relationships betweenkeywords and data items. As a result, current P2P keywordsearch systems supports only naive text-match search andcan find only data items with a keyword (or meta-data) exactlyindicated in a query.In this paper, we show an efficient P2P search systemwhich increases possibility of discovering desired dataitems. The key mechanism is query expansion, where a receivedquery is expanded based on keyword relationshipsmanaged in a distributed fashion by participating nodes.Keyword relationships are improved through search and retrievalprocesses and each relashionship is shared amongnodes holding similar data items. We also present implementationof our P2P search system.