Research On Recommendation Method Based On Network Consumption Psychology

The continuous development and progress of information technology have promoted the growth of online shopping industry with the Internet as the channel.At the same time,the increasing expansion of commodity types and numbers of ecommerce sites has caused the problem of "information overload".The related research thinks that recommendation system can collect user behavior information and recommend user commodity more active and intelligent to satisfy their needs and interests,so it can solve the problem of information overload effectively.With the further research in the related fields of recommendation system,excellent e-commerce recommendation system not only requires its good recommendation accuracy,but also helps users broaden their horizons,improving user experience and further improving the profit of e-commerce platform.And the way to build a user model can show the user's real needs and willingness to help the E-commerce recommendation system to improve the quality of the recommendation.With the improvement of consumers' living standard and consumption ability,consumers not only pay attention to the use value of commodities,but also pay more attention to the satisfaction of commodities to their internal psychological needs.Nowadays,how to make personalized marketing plan based on different users' consumption behavior and psychological characteristics is an important development problem for electric business enterprises.Considering that users' consumption psychology is the importance factor of their consumption decisions,this paper chooses a more humanistic perspective based on the existing consumer psychology research.The user network consumption psychology is introduced into e-commerce recommendation,and then the user network consumption psychological model is established.Combined with the existing recommendation technology,the recommendation method based on user network consumption psychology is studied in the local group buying network environment.First of all,it is considered that the characteristics or preference models of users are often built in the research of e-commerce recommendation system to analyze their rules.Therefore,this paper analyzes the relationship between commodity attributes and users' consumption psychology from the perspective of network consumption psychology,and then establishes user network consumption psychological model by using bayesian network technology.Secondly,according to the user network consumption psychology model,a bayesian network inference technology is applied to propose a recommendation method based on network consumption psychology model for the single user.Thirdly,a recommendation method based on the single user network consumption psychology model is improved.Combined with the clustering technology,a group recommendation method based on the core user's network consumption psychology model is proposed.Finally,the experiment is designed to verify and analyze the accuracy and user satisfaction of the recommendation methods proposed in this paper.