Abstract: This paper analyzes existing reputation models, proposes a P2P reputation model using trust mechanisms and recommendation mechanisms, and optimizes the model by the mentality of decision tree. It puts forward a method that can access global trust value in distributed P2P system, and resolves signal peer invalid problem. Experimental results prove that this model is better than existing models in estimating trust value accuracy and distinguishing untrue peers.

Key words: P2P, reputation model, trust value, recommendation

摘要： 分析现有信誉模型，提出一种使用信任机制和推荐机制的P2P信誉模型，利用决策树思想优化该模型。给出一种在分布式P2P系统中存取全局信任值的方法，解决了单点失效问题。实验结果表明，该模型对信任值的估算准确度以及判别虚假节点的能力高于现有模型。

关键词: 对等网, 信誉模型, 信任值, 推荐

CLC Number: 

• TP309