Sciweavers

Share
DASFAA
2006
IEEE

KEYNOTE: Keyword Search by Node Selection for Text Retrieval on DHT-Based P2P Networks

12 years 8 months ago
KEYNOTE: Keyword Search by Node Selection for Text Retrieval on DHT-Based P2P Networks
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 cost. 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 could 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.
Zheng Zhang, Shuigeng Zhou, Weining Qian, Aoying Z
Added 10 Jun 2010
Updated 10 Jun 2010
Type Conference
Year 2006
Where DASFAA
Authors Zheng Zhang, Shuigeng Zhou, Weining Qian, Aoying Zhou
Comments (0)
books