Application Of The Trust Mechanism In Recommendation Algorithm Of Electronic Commerce

For the current e-commerce system,its own in the process of constant development,in order to achieve better sales,goods automatic recommendation technology has been widely used in the business system,as for the current recommendation technology,there remains three main problems as followed:one is the recommendation mechanism,two is how to recommend items to meet needs of numerous user,three is how to recommend more accurately and stably.Many scholars have studied how to better optimize the recommendation mechanism.One of the methods is to introduce the trust mechanism into the recommendation technology.This paper focuses on how to introduce trust mechanism into today's e-commerce system by qualitative and quantitative methods.The research contents mainly include the following:1.Study the influence of trust mechanism on recommendation technology.This paper mainly studies the existing problems in the current collaborative filtering recommendation technology,and proposes that adding the trust relationship between users can have a positive impact on the current recommendation technology.2.Make use of the concept of community.The concept of community mainly refers to the fact that individual distribution mainly exists in the form of group,and members of groups are more closely related to each other than members outside the group.Through the purchase records of the mall to divide the community groups,research and explore the trust between users and the degree of method.3.Use quantitative method to model the trust relationship between users,and compare the traditional recommendation technology without adding trust relationship with the recommendation technology with adding trust mechanism.The trust relationship between refereed users is usually presented in detail by the trust module,and the traditional personal similarity recommendation model is fully combined.The construction of the user's personal trust model can not only greatly increase its own effective information,and update the activities of the user with close relationship timelyAccording to relevant data and experiments,when the personal trust mechanism is perfectly combined with the recommendation system of e-commerce,it can solve the shortcomings of sparse data and cold start in some recommendation algorithms and improve the accuracy of commodity recommendation compared with the traditional collaborative filtering recommendation system.