Abstract: Maintenance of semantic information and selection of intelligence route in the P2P semantic overlay network are challenging problems. According to the small world theory in network, a new P2P semantic route model with classified nodes was proposed. First, node ontology was built in order to describe the network structure and the information in the node. Then, the formation of message routing and the method of nodes classification were created. At last, the algorithm of similar knowledge Searching and the algorithms of message routing were imported to support semantic information querying and routing. It is shown, through experimental evaluation, that the model and its algorithm can efficiently support the semantic searching functionalities in the P2P network.

Key words: Semantic Route, P2P, Classified Nodes, User Preference

摘要： 对于P2P语义覆盖网络，语义信息的维护和智能路径的选择是实现的难点。根据小世界原理，提出了一种新的基于节点分类划分的P2P语义路由模型。通过建立节点本体来描述节点的网络结构和节点下的内容项，在此基础上创建了路由消息格式和节点分类划分的方法，然后创建了支持内容语义查询的节点内相关性内容查询算法和节点间消息路由算法。通过实验对比，该语义路由模型能够提高P2P系统下的内容查找速度并且能够显著降低占用的网络带宽。

关键词: 语义路由, P2P, 网络节点分类, 用户偏好