Study On The Trust And Reputation Model In The Service-oriented Online Environment

With the development of the international information technologies, Gradually, the environment of the traditional business start to change, it from the closed competition technology turns into an open, shared diverse online service-oriented environment, more and more clients takes part into it, and enjoy the benefit from the convenience and efficient internet technology services. However, because of the lack of trust mechanism, it appears a large number of illegal business behavior, such as fraudulent online transactions, Internet fraud, it has blocked the healthy development of it. Encryption, identity verification these kind of traditional network security technologies cannot solve the problems of mutual trusts, so in the online service-oriented environment, we must adopt trust mechanisms to maintain trading order.Trust and reputation is one kind of cross-subjects that combines sociology, psychology, economics, management science and other subjects related to that field. This paper starts from the perspective of computer science,it combines with many types of subjects, also analyzes the current research and background of the trust reputation. Meanwhile it carries out the classification research on trust, it analyzes the nature of dynamic, implicitness, asymmetry, transitiveness, asynchrony, gravity as well as reputation,the extension of trust.Based on the summary of a variety of models, this paper combines with present application of online reputation system, and analyzes the key issues needs to be addressed during the design period, it including the long-live identity, rating mechanism, incentive mechanism, malicious attack, reputation updating method, ect.This paper is based on the Sporas model, carries out a kind of online service-oriented environment SOETrust reputation model, aiming at the lack of rating mechanism, and it puts forward a refined QoS rating system to provide an accurate way for served agent to measure the QoS of service agent. Aggregation algorithms of reputation combines with the background characteristics of online services, such as service value, rating agent reputation, and penalty factor to punish and shock fraudulent agent.