论文部分内容阅读
针对使用分类层次来描述数据语义的分布式数据源,提出了一种支持数据共享的基于super-peer的语义覆盖网络SSON.SSON能够根据数据的语义,动态地将peer划分成多个语义簇,语义簇之间组织成语义覆盖网络.每个语义簇由一个super-peer和一组peer组成,仅负责回答其语义子空间上的查询.查询首先根据其语义被路由到适合的语义簇中,然后被转发给包含结果的peer.同时给出了相关的算法,并进行了实验研究,实验结果表明,SSON具有良好的可扩展性,并在查找性能和代价之间取得了一个良好的折中.
Aiming at the distributed data sources that describe the semantics of the data by using the classification level, a super-peer-based semantic overlay network SSON supporting data sharing is proposed. According to the semantics of the data, the peer can be dynamically divided into multiple semantic clusters, Semantic clusters are organized into semantic networks, each composed of a super-peer and a group of peers, which is only responsible for answering queries on its semantic subspace. The query is first routed to suitable semantic clusters according to its semantics, Then it is forwarded to the peers that contain the result. At the same time, some related algorithms are given and experimental studies are carried out. The experimental results show that SSON has good scalability and a good compromise between search performance and cost .