P2P环境下基于推荐的信誉计算方法

A reputation evaluation algorithm based on recommendation in P2P environment

通过节点间信誉评价的相似度和信息熵的计算,提高了推荐信誉的可用性,降低了共谋推荐行为对间接信誉合成的消极影响。通过Einstein算子将多个推荐信誉合成间接信誉,体现了信誉评估的模糊性的特点。实验结果表明,该方法可以增强推荐信誉的可用性,在一定程度上减少共谋推荐对信誉真实性造成的影响。

Through computing the semantic relevancy of reputation evaluation among peers and the information entropy, the availability of recommendation reputation is improved, and the negative effect that the collusive recommendation has on the indirect reputation composition is reduced. The aggregation of recommenders＇reputation to an object is by Einstein arithmetic operators, which reflects the fuzzy characteristic of reputation. The experiment results show that the method could enhance the availability of recommendation reputation, and reduce the influence of the collusive recommending on the authenticity of reputation to some extent.