Grid resource discovery based on semantic P2P communities
Juan Li, Son T. Vuong · 2006
Grid technologies enable the sharing of a wide variety of resources. The full use of these resources requires effective resource discovery mechanisms. However, the complicated and dynamic characteristics of grid resources make sharing and discovering them a challenge. In this paper, we propose a semantic community approach to enable efficient resource discovery in grids. The system clusters nodes into communities according to their semantic properties. The community construction and maintenance is fully decentralized and self-organizing. This structure helps prune the searching space and reduce the cost of searching. The system exhibits many desirable properties: it supports complex queries and is fully decentralized, scalable, and efficient. Our simulation results show how searching the grids can take advantage of semantic communities to reduce searching costs and improve the quality of results.