Research And Implementation Of Auction System Based On Trust Management

Online auction has the characteristics of anonymity,dynamics and randomness.Traders exchange information only through virtual networks,lacking trust foundation,which may lead to greater risks in transactions.This paper proposes a new trust mechanism for online auction,which is based on the proposed trust mechanism,improves the trust influencing factors,adds transaction amount,transaction time,evaluation credibility factors,and integrates reward and punishment mechanism.The trust degree of merchants can be obtained through the reputation-based online auction trust mechanism proposed in this paper.Users can choose whether to trade with merchants according to the trust degree of merchants.In this paper,the existing trust model is selected as a simulation comparison to verify that the trust mechanism in this paper can effectively prevent malicious acts such as collusion.At the same time,network information is numerous,and some network users are often troubled by complicated and redundant information.The online auction system based on trust management can provide personalized recommendation services for users.Among them,collaborative filtering recommendation is the most popular recommendation technology.However,the traditional collaborative filtering recommendation has the disadvantages of cold start,sparse data,expansibility and support attack.This paper improves the existing collaborative filtering recommendation algorithm and implements a collaborative filtering recommendation combining trust.Combined with trust collaborative filtering recommendation algorithm,the project scoring matrix of bidding users needs to be established first,then the similarity and trust degree are combined to find the user's nearest neighbor,to achieve predictive scoring for users' unrated projects,and then to provide accurate recommendations for users according to the predicted scoring level.Finally,the accuracy of collaborative filtering recommendation algorithm combined with trust is measured by calculating MAE value.