开放式网络信任评价研究

STUDY ON TRUST EVALUATION FOR OPEN NETWORK

为了解决开放式网络的安全性问题,提高实体间交互的可靠性及成功率,考虑信任的模糊性与随机性本质,提出了基于云模型理论的信任评价模型,实现信任定性描述与定量度量的统一。为了提高推荐的可靠性,引进信任的不确定度计算,并通过信任惩罚算法,防止交易实体的恶意行为。为保证信任评价结果的客观性与可信性,采用加权的信任云合并算法进行信任综合,并通过信任云的相似度计算算法,实现信任的决策。仿真实验分析验证了有效性和合理性。

To resolve security issue of open networks and improve reliability and success rate of the interaction between entities,weconsider the fuzziness and randomness natures of trust,and propose a cloud model theory-based trust evaluation model,which realises theunification of qualitative description and quantitative metric of trust.In order to improve recommendation reliability,we introduce uncertaintycalculation of trust,and the malicious behaviours of trading entities are prevented by trust punishment algorithm.In order to guarantee theobjectivity and reliability of evaluation result,we use the weighted trust cloud merging algorithm to carry out trust integration,and realisedecision-making of trust through similarity calculation algorithm of trust cloud.The validity and rationality of this approach are analysed andverified by simulation experiments.