Product Personalized Recommendation Method And Svstem Based On Used Car Trading Platform

In recent years,with the increasing number and variety of auto products on the used car trading platform,there has been asymmetric information between buyers and sellers on the platform.At the same time,it is difficult to provide users with personalized demand services using traditional search technology,which restricts the further development of the platform.To this end,this article proposes a solution for applying a personalized recommendation system in a used car trading platform.And take this as the research object,build a collaborative filtering personalized recommendation algorithm based on the change of user interest.And based on Hadoop technology as a basic architecture,a personalized recommendation system for second-hand car trading platform products is designed to provide users with more convenient,reliable,and efficient car product push services.The paper work is summarized as follows:(1)Introduces the research background and purpose of personalized recommendation technology for second-hand car trading platform,and expounds the development status and existing problems of second-hand car trading platform.Aiming at the problem of vehicle label semantic ambiguity in second-hand car trading platforms,a solution for constructing complex label networks and clustering them to generate different label families is proposed.(2)Aiming at the problems of changing user interest of second-hand car trading platforms and sparse matrix data,a personalized recommendation algorithm for collaborative filtering based on user interest changes was constructed.The algorithm introduces the time factor in constructing the user interest model.By combining the label weight and the time model to achieve the dynamic update of the user interest model,it effectively solves the problem of constantly changing user interest in the used car trading platform and matches the user's real-time recommendation.Car product information.By establishing a user-interest topic interest degree relationship matrix instead of a scoring matrix to calculate the similarity between users,it not only reduces the dimension of the scoring matrix,but also solves the problem of sparse matrix data in the used car trading platform,thereby improving the recommendation of cars for users Product information efficiency.Through design experiments,the accuracy and recall rate of the recommendation algorithm and other recommendation algorithms constructed in this paper are compared and analyzed to verify the superiority of the proposed algorithm in terms of recommendation accuracy.(3)Design personalized recommendation system for used car trading platform products.The requirements of the platform are described in detail,and the recommendation system design is completed using distributed processing technology based on Hadoop big data.The recommendation system is divided into a database collection and storage module,a recommendation engine module,and a system management module.The recommendation engine module is the core module of the personalized recommendation system for second-hand car trading platform products,and discusses the implementation mechanism of the module recommendation in detail.The recommendation system enables the used car trading platform to have a "thousands of people" service capability,and better provide users with intelligent car product information push services.