Research On The Recommendation Algorithm Based On Trust Relationship And Preference

With the advent of service-based computing,a large number of cross-domain E-services have emerged.Users can easily select products.Platform websites can also stimulate data traffic and occupy a larger market share.However,the increase in the number of online products has also caused difficulties for users to select desirable products from a large number of products.Under such circumstances,the introduction of an excellent personalized recommendation system is very helpful and can improve this situation.An efficient and robust recommendation algorithm is the key to the personalized recommendation system for accurate recommendation.The traditional collaborative filtering recommendation algorithm has several problems such as cold start,sparse data and weak anti-attack ability,which will affect the availability and accuracy of the recommended algorithm.In order to solve these problems,the researchers proposed the trust-based recommendation system TARS(Trustaware recommendation system)and presented a large number of related algorithms.These algorithms rely on the trust relationship between users to provide the degree of trust,which is an important measure of relevance other than preference relevance.It can alleviate the problems of traditional collaborative filtering algorithms to a certain extent and improve the recommendation effect,but there is still room for improvement.This article aims to propose a recommendation algorithm with more accurate recommendation results.The article begins with a detailed description of trust and highlights the current popular trust model.At the same time,the article introduces the classic algorithms including traditional collaborative filtering and trust recommendation.Then the comparative analysis between these algorithms is carried out.In this context,in order to obtain better recommendation results and greatly increase the coverage of algorithms,this paper proposes a collaborative filtering algorithm based on trust and preferences,TPCF(Collaborative Filtering Algorithm Based on Preferences and Trust).On the one hand,the algorithm introduces trust transfer in social networks and utilizes it to establish a trust transfer model to obtain trust among users.On the other hand,based on the score data,the similarity between users in the system is calculated.On the basis of similarity between users' trust and preference,according to the characteristics of social networks,users' trust and preference are dynamically combined to obtain comprehensive recommendation weights.The comprehensive recommendation weights can replace the traditional similarity measurement standards for user-based collaborative filtering recommendation.Besides,the attenuation of the user's scoring weight with time is taken into account to make the prediction results more accurate.The TPCF algorithm combines the advantages of traditional collaborative filtering algorithms and trust-based recommendation algorithms to further improve the effectiveness and usability of recommendation.Numerous experiments have been carried out to analyze the influence of input parameters,convergence,and trust distribution of the trust propagation model.Then the influence of parameter in the TPCF algorithm is studied.Finally,the TPCF algorithm,collaborative filtering algorithm,trust-based algorithm and other existing trustbased recommendation algorithms are compared in terms of accuracy,coverage,availability,and attack resistance ability.The experiment results show that TPCF recommendation algorithm can overcome the problems of traditional collaborative recommendation algorithm and obtain better recommendation effect to some extent.