International Journal of Advanced Research in Computer Engineering & Technology
Online reputation systems are playing increasingly important roles in influencing people's online purchasing/downloading decisions. In this paper, the authors propose a scheme user correlation algorithm. The existing mechanisms time domain anomaly detection and trust model based on dempster safer theory not provide better efficiency. First, limit the maximum number of ratings each user could provide within the certain time duration. Second investigating users rating behaviors by building user trust, such as a personalized trust model, which evaluates a user's judging power as the inverse of this user's rating variance.