Dynamic Recommendation Trust Evaluation Model Based on Mobile E-Commerce.

Gang Wang,Xiaolin Gui
DOI: https://doi.org/10.1109/hpcc.2012.189
2012-01-01
Abstract:In the past trust model, fraud and cooperation cheating of malicious nodes were restrained by reducing the trust value of malicious nodes, but few recommendation nodes for cooperation cheating have be punished. Due to lack to punish to malicious recommendation of recommendation nodes, enormous malicious recommendation nodes remain in network and search next criminal opportunity constantly. This paper proposes a dynamic recommendation trust evaluation model based on mobile e-commerce environment. By using bidding strategy and incentives mechanism, let recommendation nodes recommend actively and maintain competition, and nodes could gain corresponding rewards or punishments at the same time. In addition, we propose a novel evaluation method for quality of recommendation service, and the method overcomes effectively monotony and inaccuracy of the past evaluation means, and solves a problem that recommendation of the higher trust node is the more credibility than the others. From simulation experiment analysis, we find the trust model shows good restraint to malicious recommendation and collaborative cheating attack of malicious nodes.
What problem does this paper attempt to address?