摘要
声誉管理是在交易伙伴之间建立信任的有效手段.针对现行电子商务声誉系统的不足,在分析相关研究的基础上提出了一种新的声誉系统模型.模型采用连续数据表达声誉,在支持不实声誉评价识别与过滤的同时,将交易价格、交易时间等交易上下文因素也考虑进来,增强了模型的适应性和动态性.另外,引入了鼓励客观声誉评价的机制.模型在设计上注重可操作性,是一种轻量级声誉系统.仿真实验说明了模型及相关算法的有效性.
Reputation management is effective method to build trust between the buyers and sellers. Aiming at the flaw of the existing reputation system, based on analysis of related work, a new reputation system model for electronic commerce is provided. Real number is used to express and compute reputation. Filter mechanism is introduced to find out collusion cheat, at the same time, some context factors, such as transaction price and time, are introduced to adapt the dynamics of the reputation. Actual and objective reputation opinion is encouraged. This reputation system model is a lightweight and easy to operation. The simulation experiment shows the trust model is effective.
出处
《小型微型计算机系统》
CSCD
北大核心
2011年第7期1365-1369,共5页
Journal of Chinese Computer Systems
基金
上海高校选拔培养优秀青年教师科研专项基金项目(slx08017)资助
上海市教育委员会科研创新项目(10YZ193)资助
关键词
声誉
信任
电子商务
联合欺诈
上下文
reputation
trust
electronic ecommerce
collusion cheat
context
