KEYNOTE: Keyword search by node selection for text retrieval on DHT-based P2P networks

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Efficient full-text keyword search remains a challenging problem in P2P systems. Most of the traditional keyword search systems on DHT overlay networks perform the join operation of keywords at document level, which consumes a huge amount of storage and bandwidth. In this paper, we present KEYNOTE, a novel keyword search system that performs the join operation at node level. Compared to the traditional keyword search systems on DHTs, KEYNOTE can greatly reduce the storage and communication cost. To forward a query to the relevant nodes for searching documents, two effective node selection methods are presented. To address the hot spot problem in Chord overlay networks, an efficient load balancing scheme is introduced. Simulated experimental evaluation with up to 8,000 nodes and over 600,000 real-world documents validates the practicality of the proposed system.

Original languageEnglish
Title of host publicationDatabase Systems for Advanced Applications - 11th International Conference, DASFAA 2006, Proceedings
PublisherSpringer Verlag
Pages797-806
Number of pages10
ISBN (Print)3540333371, 9783540333371
DOIs
StatePublished - 2006
Externally publishedYes
Event11th International Conference on Database Systems for Advanced Applications, DASFAA 2006 - Singapore, Singapore
Duration: 12 Apr 200615 Apr 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3882 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference11th International Conference on Database Systems for Advanced Applications, DASFAA 2006
Country/TerritorySingapore
CitySingapore
Period12/04/0615/04/06

Fingerprint

Dive into the research topics of 'KEYNOTE: Keyword search by node selection for text retrieval on DHT-based P2P networks'. Together they form a unique fingerprint.

Cite this