A Study Of Readers' Borrowing Behavior Of The Public Library Oriented To Two-way Book Recommendation

With the rapid development of information technology,readers have access to a wider range of information channels,and public libraries are facing transformation.In the huge flow of information,how to effectively improve the utilization rate of university libraries' collections and adapt to the needs of the majority of readers while enhancing the core competence of public libraries themselves becomes the focus.With many readers in public libraries,there are a large number of reader and book records in the management system,and these data are valuable for analysis in many aspects.In this paper,we take the role of a third party,combine the dual perspectives of readers and books to study and analyze the data of readers' borrowing behavior in a public library,use the Clustering Analysis and Association Rules in machine learning to establish a two-way recommendation model for public libraries,and propose the collection of public libraries The study also suggests how to build resources and make better recommendations to readers.The main findings of this study are three main points.First,readers' borrowing behavior preferences are influenced by the dual characteristics of readers' age,gender and the time of book publication,publisher and author.Second,Many public libraries do not have a high overall usage rate and the number of frequent readers is low,and the number of books with high circulation is also small,which tends to cause a waste of resources in the library.Third,readers' tendency to borrow varies widely.For example,children have a wide range of interests and tend to borrow books of different categories,which meet the requirements of children's curiosity and overall development,while adults' tendency to borrow is more homogeneous.The tendency of male and female readers' borrowing behavior is more obvious,with men preferring to borrow history and geography books and women preferring to borrow literature books.Based on the above findings,different suggestions are given for library readers' recommendation and book acquisition.On the one hand,libraries need to improve the accuracy of reading recommendation,improve the comprehensive quality of librarians,and take different recommendation methods for different reader groups.On the other hand,in the construction of collection resources,libraries should combine readers' needs,adopt various recommendation methods,and focus on building books with high demand quantity to meet the needs of more readers.