Grid computing provides key infrastructure for distributed problem solving in dynamic virtual organizations. It has been adopted by many scientific projects, and industrial interest is rising rapidly. Compared with traditional single computer system, effective service locating in Grid is difficult because of huge amount and wide-area distribution of dynamical services. This paper presents a new Grid framework that discovery services through the overlay network based on Grid service description semantic vectors generated by Latent Semantic Indexing (LSI) [1]. This framework uses WSRF as its base architecture and regards each Grid node as a register node to publish service that enables it has better scalability. Comparing with other Grid services discovery techniques those are based on simple keyword matching, our framework has better accuracy for it considers the advanced relevance among Grid descriptions.
[1]
Mark Handley,et al.
A scalable content-addressable network
,
2001,
SIGCOMM '01.
[2]
Richard A. Harshman,et al.
Indexing by Latent Semantic Analysis
,
1990,
J. Am. Soc. Inf. Sci..
[3]
Peter Steenkiste,et al.
Darwin: customizable resource management for value-added network services
,
2001
.
[4]
Donald F. Ferguson,et al.
The WS-Resource Framework
,
2004
.
[5]
I. Foster,et al.
The Physiology of the Grid
,
2003
.
[6]
Elizabeth R. Jessup,et al.
Matrices, Vector Spaces, and Information Retrieval
,
1999,
SIAM Rev..
[7]
Steven Tuecke,et al.
The Physiology of the Grid An Open Grid Services Architecture for Distributed Systems Integration
,
2002
.
[8]
Jim Waldo,et al.
The Jini Specification
,
1999
.