一种P2P网络的信息优化检索算法的仿真分析  

A P2P Network Information Retrieval Algorithm Optimization Simulation Analysis

作　　者：陈春燕[1] 

机构地区：[1]北京信息职业技术学院,北京100018

出　　处：《科学技术与工程》2013年第9期2572-2578,共7页Science Technology and Engineering

摘　　要：为了解决P2P网络系统信息检索效率低、不能有效解决跨文本搜索,提出节点兴趣域聚类和信息量声誉激励的P2P网络检索机制。在该机制中,首先对网络系统中节点持有的数据信息量进行基于相似度和兴趣度阈值的兴趣域聚类;然后依据节点数据信息的声誉激励策略对兴趣相邻节点进行兴趣树构造,同时对用户输入的搜索关键字进行语义分析和个性化的辅助语义选择。将与查询信息向量最接近的节点持有信息量返回给用户,并对该数据信息量进行声誉激励评价和更新。实验仿真证明,该算法基于兴趣树的动态构造,能够避免结构化P2P网络系统对中心节点的过度依赖;同时检索的向量是基于用户个性化辅助语义生成的,能够有效地提高查询率和查准率。In order to solve the P2P network information retrieval efficiency low, cannot effectively solve the cross text search, node interest domain clustering and information reputation incentive P2P network retrieval mecha- nism are put forward. In this mechanism, first of all, the data information which holds on the node of network sys- tem cluster with interest domain based on similarity and interest degree threshold, then, interest tree structure to in- terest adjacent nodes through the node data information reputation incentive strategy, the semantic analysis and per- sonalized auxiliary semantic choice for users to enter search keywords, returned to the user which hold node infor- mation and query information vector closest to the data and information for reputation incentive evaluation and up- date. The simulation results show that the algorithm can avoid structured P2P network system for center node excessive dependence based on the interest of the dynamic tree structure, and at the same time, retrieval vector is based on user personalized auxiliary semantic formation, can effectively improve the inquires the ratio and the precision ratio.

关 键 词：P2P网络 节点兴趣域 声誉激励 信息相似度 

分 类 号：TP393.3[自动化与计算机技术—计算机应用技术]