Research And Implementation Of Personalized Information Recommendation System For"Energy Headline"

With the rapid development of Internet,the amount of data is increasing,which brings the problem of information overload.Facing the problem of it,the client is facing how to provide satisfactory for different user experience of reading,users are facing how to select their love information in order to improving the efficiency of reading.In recent years,the research of personalized recommendation system has gradually become the focus of all walks of life,to provide users with their own personalized recommendation is an urgent problem to be solved.Personalized information recommendation system can accurately mining user preferences by using the recommended strategies,and improve the user satisfaction.At present,information recommendation system on the market mostly through the level of heat press make recommendations for the user,or the user data according to the reading of history,through the recommendation strategy based on content recommendation and preference consistent information,unable to find new user preference.Personalized Information Recommendation System for“Energy Headline” wants to create a private ordering APP,The purpose of this study is to find an efficient recommendation strategy.According to the characteristics of different user types and behavior,to help users find the new point of interest,quickly and accurately recommend high satisfaction the information for the user.The main content of this paper includes,by the studing of information topic model and user interest model proposed a topic model based personalized recommendation strategy,to solve the cold start problems and data sparseness problem often exist in recommender systems,the modeling method adopted LDA theme model in this paper to establish information model,and its application to the user interest model,in order to improve the accuracy of recommendation,joined the scene information in the model,to solve the problem of time and location factors on the recommendation result.The design of the recommended scheme,the system for new users recommend popular information to solve the problem of the cold start,when the user reading data exceeds a certain number,a method to solve the data sparseness problem in collaborative filtering algorithm is to calculate the similarity between users by similarity of user topic interest,Finally,through the experimental analysis,the model and the recommended scheme can achieve the purpose of accurate recommendation.