A robust trust management model based on recommendation was proposed to deal with three kinds of common attacks in E-commerce system including self-promoting attack, slandering attack and whitewashing attack. This paper introduced some new concepts such as the time decay function based on the forgetting curve, consumer behavior similarity and dynamic confidence for direct trust value etc. It thought about some factors affecting trust evaluation containing the time factor, recommendation trust factor, number of transactions, trading satisfaction and etc. Furthermore, the model also solved the cold start problem in reputation system by setting the initial trust constant between customers. On the other hand, this paper designed a mechanism to detect the inconsistent feedback from normal buyers. For the simulation experiments, the real transaction data was used, and the results verified the effectiveness of the reputation model. It had been proved that the model proposed in this paper had more advantages than the existing reputation models for E-commerce system.
[1]
Cristina Nita-Rotaru,et al.
A survey of attack and defense techniques for reputation systems
,
2009,
CSUR.
[2]
Han Yi,et al.
Trust and e-commerce: a study of consumer perceptions
,
2003,
Electron. Commer. Res. Appl..
[3]
Ling Liu,et al.
A reputation-based trust model for peer-to-peer ecommerce communities
,
2003,
EC.
[4]
A. Jøsang,et al.
Challenges for Robust Trust and Reputation Systems
,
2009
.
[5]
Chunxiao Xing,et al.
Similarity measure and instance selection for collaborative filtering
,
2003,
WWW '03.
[6]
Ernesto Damiani,et al.
Fuzzy techniques for trust and reputation management in anonymous peer-to-peer systems
,
2006,
J. Assoc. Inf. Sci. Technol..