【摘要】 针对信任评估的主观性、模糊性以及无法精确地描述和验证等问题,提出信誉评估算法。算法定义了信任度、隶属度、信任关系、关键属性以及模糊变换等基本概念和关键属性集U、代理集Y、评价集V、评价集矩阵R四个要素。根据模糊数学理论,将代理X的信任程度以及各代理对被代理的各关键属性的评估进行模糊变换,得到即将交易的代理Y的信任向量,并根据实际对原代理X的各信任值向量化,最终得到对代理X的信任值或信誉。实例表明,算法能用直观、简洁的语义定量地描述信任值,反映了实体之间信任的模糊性,解决了关于模糊性实体之间的信任建模问题,适用于电子商务、电子政务等领域。更多还原

【Abstract】 Because trust itself had objective and fuzzy,and could not descript and validate it exactly,this paper proposed a reputation evaluation algorithm based on fuzzy logic.Algorithm defined some basic concepts such as trust degree,trust relation,key attributes and fuzzy transformation and four elements such as key attributes set U,agents set Y,evaluation set V and evaluation matrix R.According to fuzzy transformation in fuzzy map logic,it could get the trust vector of the given agent through the fuzzy transformation between trust degree of agent X and the evaluation of key attributes of the given agent.If necessary it could vectorize the trust levels,then could get trust value of the given agent X.From the example,algorithm not only reflect the fuzzy of trust but also a visual and simple quantification description mechanism.It can apply in e-commerce,e-government and so on.更多还原