Research On Core Book Recommendation Model Of University Library Based On R Index

According to the statistics of the Library and Information Work Steering Committee of the Ministry of Education of the People's Republic of China,the purchase of paper literature resources in China's university libraries has always remained at a relatively high level.With the continuous rapid growth of paper resources,the paper literature utilization rate has shown the downward trend,how to improve the utilization rate of library paper literature resources,and give full play to the value of paper literature resources has become the focus of university libraries.At the same time,in the face of a large amount of paper literature resources,how to find the literature resources that you need conveniently and efficiently has become a problem for library users.It can be said that the contradiction between the low utilization rate of paper literature resources and the difficulty for library users to find the required resources conveniently and efficiently is increasingly apparent.Book recommendation is an effective way to solve this problem.From the literature research situation,resource recommendation based on collaborative filtering is the mainstream mode of book recommendation system,but collaborative filtering recommendation has problems such as data sparsity and scalability,which leads to poor performance in practice.At present,the domestic university library book retrieval system is still based on the traditional TOP-N popular book recommendation,which is difficult to meet the needs of users.Based on this,this study proposes a core book recommendation model based on R index,which is divided into two groups according to the students' group in the recommendation system,and then according to the students' profession,gender,grade,target book categories and other attributes for further subdivision,and finally according to the student group to generate a list of recommended books,the core link is to use the R index to identify the core books in the previous borrowing records.For the non-login users of the book recommendation system,because they cannot get the personalized attribute information of their profession,gender,grade,etc.,the book recommendation will first use the book coupling relationship analysis to find the most closely related books,and then use the R index to identify the core books and generate recommended book list.Because the model divides the user into groups,and tries to pre-build the core library through the R index,which is the recommended bibliography,it changes the traditional TOP-N popular book recommendation model in the past,making the process of recommending the book more scientific and reasonable.The recommendation model attempts to comprehensively consider the system login and non-login users,making the recommendation system more universal,and providing a new idea for university library book recommendation.The first chapter of this research is the introduction,systematically combing the research background and significance,research content and methods,research innovations,etc.The second chapter is the relevant theories and research review,the relevant research on the recommendation system and the research review of the book recommendation.The third chapter analyzes the borrowing characteristics of different user groups,and analyzes the borrowing characteristics of different user groups from the perspectives of profession,gender and grade.The fourth chapter is the study of book coupling relationship,using Social Network Analysis method to explore the internal relationship between books and find the most closely related books.The fifth chapter is the research on the construction of the R index core library,it introduces the feasibility of applying the R index to the field of borrowing data analysis,and uses the R index to identify the core books.The six chapters are the recommended model and applied research,introduces the model recommendation process,and tests the model through examples.The seventh chapter is the conclusion and outlook,it combs the research process,research results and main problems in detail,and attempts to propose future research directions.