Research On Personalized Advertising Recommendation Based On User Model

With the rapid popularization of Internet and the promotion of smart carrying products,more and more people are investing in the Internet,and the Internet has become a new carrier of modern advertising business.Internet advertising has high efficiency,wide coverage and so on,so it has received unprecedented attention.The traditional Internet advertising is full of randomness and uncertainty,lead users faced with overwhelming advertising in the Internet,reducing the user's online experience,resulting in low conversion rate of advertising.In order to improve the user experience,researchers have proposed personalized advertising recommendation technology,and become the research focus in recent years.The core of personalized advertising recommendation technology is the user model,and the quality of personalized advertising service is directly related to the accuracy of the user model.This paper proposes a personalized advertisement recommendation technique based on explicit implicit information model.The following research work has been carried out:(1)An improved implicit modeling method is proposed.The traditional modeling method of implicit user will log all information as the modeling information,but the user's interest is the real time change of the implicit improved modeling method adopting the method of historical information classification,only related to the query history information as the user modeling document set.(2)A user modeling technique combining explicit and implicit information is proposed.Explicit user modeling with implicit information according to the user information to initialize the user model submitted by the user,and then through the Internet history information for users is analyzed,the construction of implicit user model,the initial user model update.(3)A collaborative filtering advertising recommendation algorithm based on user model is proposed.Collaborative filtering technology is based on the user project scoring matrix to mine the user similar set,and recommend the information to the target user through the users of similar interest in thenearest neighbor set.This paper will use the user interest model presented above to collaborative filtering algorithm,using the user model matrix instead of score matrix,the experimental results show that collaborative filtering algorithm can improve the accuracy of user model based on advertising recommendation.