基于朋友聚类的语义覆盖网构建方法研究

P2P网络是一种极具潜力的信息共享方式,当前对于它的研究成为网络技术的热点问题.P2P网络的分布、分散、组织困难等特点,使P2P网络在应用中有很多不足,这就要求提高P2P网络的应用效率,优化查询过程等.以一个无结构的,可以用无向图来表示的P2P网络为基础,首先构造了一个描述节点上文件的元数据模型,在这个模型的基础上进行语义聚类,并定义了一个聚类特征向量来描述生成的聚类.在完成本地节点上的聚类以后,再进一步进行节点间的聚类,提出了朋友聚类的概念,基于朋友聚类构建语义覆盖网,它可以大大提高查询效率.

Research on Creating Semantic Overlay Networks Based on Friend Clustering

Peer-to-peer network has shown a great potential to become an excellent information sharing way,it has become the hot issue of network technology.However,the distribution,decentralization,difficult to organize and other features of P2P networks made it turn out a lot of disadvantages,so it is very necessary to improve the efficiency of applications and optimize the process of querying in P2P networks.In this paper,we began from a non-structure P2P network,proposed a metadata model to describe those documents on one peer,based on this model implement semantic clustering,and then defined clustering feature vectors to describe those clustering.After achieved the local clustering we begin to carry out clustering between peers,and then introduce the concept of friend clustering.Based on friend clustering creating semantic overlay networks,it could greatly improve the efficient of querying.