主题驱动的P2P分布式信息搜索机制研究

Peer—to—Peer（P2P）对于分布式文件共享具有很好的前景，但当前的P2P系统仍然缺乏有效的信息搜索机制．本文提出一种主题驱动的P2P信息搜索机制，通过对节点上的文档进行聚类获得全局主题，然后将包含有相似主题的节点组织到一起构成主题覆盖网络．当在P2P网络中进行信息搜索时，根据查询与主题的相关性路由查询，从而改善搜索效率．本文详细阐述了进行主题驱动搜索的索引结构、主题聚类方法、主题覆盖网络的构造与维护算法．在Chord上的模拟实验结果表明，主题驱动的P2P信息搜索机制可以减少信息搜索时的平均网络带宽和平均搜索路径长度，提高搜索的成功率．

Topic Driven Distributed Information Searching in Peer-to-Peer Networks

Peer-to-Peer (P2P) systems are emergent and promising technique for resource sharing in distributed environment, but efficient information search in P2P system is still a challenging problem. This paper presented a topic driven search mechanism to improve the search capacity of the existing structured P2P system. The documents were clustered to form topics and the nodes which include relevant topics were linked to form topic overlay network. During the search process, the query only was routed to the nodes which are relevant to the query. The index structure, the clustering method and the construct and maintenance algorithm of the topic driven search are given in detail. The simulation experiment results on the Chord showed that the topic driven search mechanism is efficient to reduce the network bandwidth and the search path length, and to improve the search precisions.