Research On Personalized Recommendation Algorithm Based On User Portrait

With the rapid development of online shopping platforms and the increasing number of shopping people online,online shopping has become one of the most important ways for readers to buy books.How to improve the user's shopping experience,attract more and more users in online shopping platform to buy books,become the biggest challenge in the development of e-commerce platform.In this paper,personalized book recommendation based on personas can recommend users' favorite books,so that users can have more choices when buying books,and choose books that are satisfactory to them.This is a detailed analysis of the various types of information that users leave in the purchase of books on the online shopping platform,and by improving the current popular content-based recommendation algorithm,the improved algorithm will be able to incorporate the user's personal information,the evaluation of the book,and the evaluation of the book,and create a model of the personas,which will ultimately be the result of the user's accurate reference to the book.The main work of this article includes:(1)Introduce the classic content-based recommendation algorithm in the field of recommendation,apply the algorithm to books to purchase this particular scene,analyze the advantages and disadvantages of the algorithm in this scenario,and propose the detailed improvement plan based on the information in the data set.(2)A detailed description of the proposed algorithm based on content,new user similarity algorithm in the modified algorithm,and book scoring prediction method,solving the problem of the new user when the new user first came to the purchase of the book on the online shopping platform and the old user's evaluation of the book.(3)Via the personas concept introduction,analysis of personas in the scenarios need to design the dimension of the books.The modeling process of the personas,the problem of how the algorithm and the data are combined are described in detail,and finally the personas model is obtained.(4)To analyze and pre-process the data set,and then to use the data set to be processed in the form of the modified and improved version of the proposed algorithm,and by contrast,Experiments show that the improved content based recommendation algorithm is higher than the recommendation algorithm based on content recommendation rate and recall rate.