A Reputation Revision Mechanism to Mitigate the Negative Effects of Misreported Ratings
International Workshop on Entertainment Computing(2015)
摘要
Reputation systems aggregate the ratings provided by buyers to gauge the reliability of sellers in e-marketplaces. The evaluation accuracy of seller reputation significantly impacts the sellers' future utility. The existence of unfair ratings is well-recognized to negatively affect the accuracy of reputation evaluation. Most of the existing approaches dealing with unfair ratings focus on filtering/discounting/aligning the possible unfair ratings caused by malicious attacks or subjective difference. However, these approaches are not effective against unfair ratings in the form of misreporting (e.g., a well-behaving buyer misjudged a seller and provided a negative rating to a transaction which deserves a positive one, and the buyer is willing to revert the misreported negative rating). In this case, how should the buyer undo the damage caused by such misreported ratings and help the seller recover utility loss? In this paper, we propose a reputation revision mechanism to mitigate the negative effects of the misreported ratings. The proposed mechanism temporarily inflates the reputation of the misjudged seller for a period of time, which allows the seller to recover his utility loss caused by the misreported ratings. Extensive realistic simulation based experiments demonstrate the necessity and effectiveness of the proposed mechanism.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络