一种基于内容相似度和推荐反馈的信任模型

Trust Model Based on Content Similarity and Recommendation Feedback

在已有P2P模型的基础上提出了基于内容相似度和推荐反馈计算节点推荐值的对等网络信用模型IPBS(Integrated-partial based si milarity Trust)。该模型利用节点间的内容相似度来评价节点提供推荐服务的能力,根据每次交易的内容不同而改变节点间相似度值;同时依据节点交易历史时间和推荐反馈值自适应动态地调整节点的推荐值;实例表明,IPBS节点间推荐值,通过参考节点内容相似度、交易历史时间和推荐反馈3种机制,加强了模型的动态适应能力和搜索服务的效率。

This paper proposed a P2P trust model-IPBS which introduces content similarity and recommendation feedback to calculate the peer＇s recommended values. The recommendation ability of nodes was evaluated according to content similarity between peers. The similarity was changed by the different transactions. The model can adaptively update the recommended values based on historical trade time and recommendation feedback. Experimental results demonstrate that the trust model has advantages of dynamic adaptability and is highly effective in the search efficiency.