关键词: super-peer, 分类层次, 语义查找, 路由算法, 语义簇

Abstract: An approach to framing the semantics-based super-peer networks is proposed aiming at the data semantic space consisting of taxonomic hierarchies. The approach can dynamically group peers into many semantic clusters based on their individual semantics of data object. Then, the semantic clusters are organized into a semantic overlay network with appropriate routing relations established between them. Each and every semantic cluster consists of a super-peer and a set of peer nodes, and only responsible for answering queries in its semantic sub-space. A query is first routed to the appropriate semantic clusters by an efficient searching algorithm, then forwarded to the specific peers that hold the relevant data objects. Experimental results indicated that the approach achieves a competitive trade-off between search efficiency and cost, thus offering higher search performance and lower search cost to enhance the network scalability.