一种基于多主题的网络模型构建算法  

Network model algorithm based on multi-subjects

作　　者：郭磊[1] 刘方爱[1] 杜玉霞[1] 姜浩然[1] 

机构地区：[1]山东师范大学信息科学与工程学院,济南250014

出　　处：《计算机应用研究》2011年第3期1111-1113,1118,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(90612003);山东省自然科学基金资助项目(Y2007G11)

摘　　要：针对P2P网络中很难获取资源全局信息的问题,提出一种基于多主题的网络模型构造算法。算法根据网络中主题分布情况,在逻辑上将网络划分为若干主题子网,使具有相似或相同主题的节点聚集在一起,从而可以在子网范围内获取某类资源的近似全局信息。实验结果表明,该算法具有较快的收敛速度和较强的自我修复与扩展能力,能够有效地进行主题子网划分和资源聚集。In P2P networks,it＇s hard to acquire global information of resources.To address the problem,this paper presented a network model algorithm based on multi-subjects.In this algorithm,divided the network into several subnets according to the distribution of subjects,and owning to the nodes that have similar subjects were congregated it could acquire resource global information.Experimental results show that the model not only can congregate nodes with similar subjects together,but also has fast convergence rate and strong ability to self-repair and expansion.

关 键 词：多主题 网络模型 超级节点 对等网络 社区 信息检索 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]