Performance evaluation of rating aggregation algorithms in reputation systems

摘要: Ratings (also known as recommendations, referrals, and feedbacks) provide an efficient effective way to build trust relationship amongst peers in open environments. The key the success of ratings is rating aggregation algorithm. Several algorithms have been proposed, however, all them are evaluated ad-hoc fashion so that it difficult compare effects these schemes. In this paper, we argue what missing evaluate different schemes same context. We first classify state-of-the-art aggregating into five categories, then comprehensively context a general decentralized inference model with respect their resistance factors, such dynamic behavior raters, dishonest ratings, on. simulation results show complicated not always good choice if take implementation cost bad raters consideration