Research On Personalized Book Recommendation Model For New Readers

With the continuous development of computer technology and network technology,contemporary people's lives,work,and learning have seen the Internet's “dashes”.In the face of massive user information,the recommendation system has been acquired in many fields such as business,social networking,and video.Widespread development.For university libraries,in the face of the ever-increasing number of books in the collection,how to quickly and efficiently obtain the required books from a large number of books,how to make full use of university library resources,these are recommended for personalized books the system generated demand,so scholars have proposed a variety of personalized book recommendation models.However,in practical applications,because most readers do not generate borrowing data in the library,it is difficult for the recommendation system to make accurate personalized book recommendations.For such new users who have not borrowed in the library,how to personalize the book recommendation,how to improve the utilization of library resources in university libraries is the main content of this study.Most of the book recommendation systems are based on the user's existing borrowing behavior data,and generate a book recommendation list for the user through a model algorithm.In order to solve the problem that the new user does not have borrowing behavior data,the paper proposes a book recommendation model based on the user's course selection data and uses collaboration.The idea of filtering algorithm is to recommend books for new users,and a rating function based on the user's loan duration is designed.At the same time,the algorithm is improved,clustered the books according to the call number,and the personalized book recommendation for new users is realized.Research contents include:Firstly,this thesis deeply studies the current situation of personalized service development in University libraries.The new users who have not borrowed books in the library are set to the target users of the book recommendation system.Then carry out the demand analysis and determine the design principles of the book recommendation model.Secondly,according to the requirements and design principles of book recommendation model for new users,the feasibility analysis of book recommendation model was carried out by studying the principle of recommendation system and classical algorithm.Then,a new user-oriented book recommendation model is constructed based on the idea of user-based collaborative filtering algorithm recommendation.Thirdly,the basic data is collected and preprocessed according to the working principle and process of the book recommendation model for new users.Then,the experimental is validated,and the influential results of different parameters in the model on the recommendation is compared and analyzed.In addition,in order to improve the recommendation effect of the model,the borrowing time function is added to the previous model to evaluate the user's preference for borrowing books.Meanwhile,in order to reduce the sparseness of the user-borrowing book matrix,the books are clustered according to the Chinese Library Classification.A new book recommendation system model for new users was conducted,and experimental verification and effect analysis were carried out.Finally,according to the recommendation results of the model,the suggestions of application of the book recommendation system model for new users in college libraries were given.