针对开放网络环境下的电子交易存在的安全性、可信度低和信任评估难的问题,在文中将灰色系统理论应用于主观信任的度量与评估,提出了一种新型主观信任评估方法,即针对一个给定实体,根据不同用户对这一实体多个关键属性的评价,经过灰色关联度分析以及灰色聚类分析综合出对于此实体的总评价。仿真实验结果表明,在评价可信度未知的情况下,该方法可以较好地保留样本的完整性,提高了信任值的准确性,能够有效地遏制虚假评价对最终评价的影响,弥补了传统方法在客观性和可信性方面的不足。
In open network, an important challenge is how to cope with security problems, low trust and evaluation problems in e- commerce. A novel subjective trust evaluation method is proposed in this paper with the employment of grey system theory. According to various evaluated trust values upon differeat properties for a given entity, a comprehensive trust value can be computed based on our method by grey correlation analysis and grey clustering analysis. Simulation results demonstrate that this evaluation method can not only keep the integrity of samples but also improves the accuracy of trust value for the entity, meanwhile, it effectively prevented the impact of baleful estimate as far as possible, and improved lack of objectivity and trust in traditional method.