摘要： 针对P2P网络中节点交易风险较大的问题，模拟社会网络的人际交互过程，提出一种基于动态推荐的信任管理模型。采用模糊聚类方法，结合交互的上下文动态地选择推荐节点，在推荐因子的计算上融入聚类分析结果，提高了推荐的可靠性。分析和模拟实验表明，该模型能有效提高推荐的准确性，增强P2P网络的可用性。

关键词: P2P网络, 动态推荐, 信任模型, 模糊聚类

Abstract: To resolve the risk problem of transaction in P2P network, by simulating interpersonal interactive process in society network, a trust management model based on dynamic recommendation is proposed. Based on fuzzy cluster algorithms and the interactive context, recommendation nodes are selected dynamically. According to results of cluster analysis, recommendation factors are computed and the reliability of recommendation is enhanced. Analysis and simulation experiments show that the method can improve the accuracy of recommendation effectively and strengthen the availability of P2P network.

Key words: P2P network, dynamic recommendation, trust model, fuzzy cluster

中图分类号: 

• TP393