Research On Service Reputation Recommending User Reputation Evaluation System

With the continuous development of Internet and e-commerce technologies,more and more users and services are involved in service computing,and it is increasingly important for users to accurately recommend services that truly meet the real needs of users from a mass of services.Service recommendations can often be explored based on service evaluation feedbacks from historical users.However,not all historical users' evaluations are trustworthy,and the trustworthiness of different users is also different.People often think that the higher the user reputation is more valuable.In the service recommendation process,the effective assessment on user reputation is attracting the attention more and more researcher.The research on service reputation-based user credibility evaluation system has gradually developed.In order to recommend new services to satisfy new users accurately and effectively,this thesis proposes a research on service reputation-based user credibility evaluation system.According to the attribute weight description given by the new user,a suitable service could be recommended by a historical user with high reputation in the system.Two kinds of feedback information are given after the new user uses the recommended service,that is,the QoS evaluation information on the recommended service and the satisfaction scoring information on the service recommendation of historical user.Based on these two types of information,trustable new users could be identified and their feedback information could be used to change the reputation score of historical users.Historical users with lower reputation could be eliminated and new historical users with high reputation could be recruited.Through the cycle of recommendation,feedback,reward,punishments and elimination and recruit,a stable and reliable historical user reputation evaluation system could be constructed to achieve higher quality service recommendations.The usability of the method is demonstrated through a case study and the validity is verified through the simulation experiment.