Research On The Personalized Recommendation Of Books Based On Collaborative Filtering

In the era of informationization,people are faced with the problem of information explosion brought about by massive information.More and more information makes it more and more time-consuming for people to obtain useful information.The recommendation algorithm is an effective solution to this problem.The success of the recommendation algorithm in the commercial field has made its application in the field of books possible.There are a wide variety of recommended algorithms,of which the most successful and most widely used is undoubtedly the collaborative filtering recommendation technology.In this paper,through the analysis and comparison of various mainstream recommendation technologies,a collaborative filtering algorithm based on graph model is selected.To solve the problem of large data volume and high matrix dimensions,an improved type of aggregation is used to reduce the dimension of the original matrix.At the same time,taking into account the problem that there is no rating information in the book borrowing data,a way of calculating the rating from the borrowing duration is proposed,and then a book recommendation model is proposed.At the same time,the feasibility of the algorithm is verified through experiments.The first chapter of this article mainly introduces the research background,the research significance and the research status at home and abroad.The second chapter mainly introduces the mainstream recommendation technology,such as content-based recommendation technology TF-IDF algorithm,focusing on a variety of collaborative filtering recommendation ideas,item-based collaborative filtering,user-based collaborative filtering and model-based collaboration filtering,meanwhile introducing various evaluation indicators of the algorithm.The third chapter introduces the idea of clustering and clustering algorithm,introduces the idea of fuzzy mean clustering algorithm in detail,and proposes an improved idea of clustering algorithm.The fourth chapter mainly proposes a new idea of recommendation algorithm.Using the idea of clustering in Chapter 3,a collaborative filtering recommendation algorithm based on graph model is improved.By considering all the items in the system as nodes in the association graph,and setting the number of users who have scored both items at the same time as their corresponding weights between the nodes,the collaborative filtering recommendation based on the graph model is realized.In chapter 5,aiming at the lack of scoring data in the book borrowing data,a scoring conversion formula based on borrowing duration is proposed,the fourth chapter of collaborative filtering algorithm is used to recommend books resources.