An Empirical Study On The Reader Borrowing Behavior And Library Resource Recommendations In University Libraries

With the development of information technology,the informatization process of universities at home and abroad has been greatly promoted.At the same time,how to improve the utilization rate of library resources in colleges and universities,meet the needs of readers,and enhance the core competitiveness of libraries have become problems that must be solved in the development of university libraries.This article establishes a recommendation model for library resources in colleges and universities from the perspective of readers' borrowing behaviors.It uses the reader borrowing behavior data of a college to conduct empirical research,and puts forward specific suggestions for college library resources in the construction of library resources.This article first discusses the research status of readers' borrowing behavior and book recommendation system both at home and abroad,compares the advantages and disadvantages of commonly used recommendation technologies,selects recommended technologies that are suitable for recommending collection resources,and establishes a collection resource recommendation model.Then from the angle of reading readers' time series and readers' role classification,the author conducts a statistical analysis on the borrowing behavior of readers,which can intuitively show the time,faculty and grade characteristics of readers borrowing books.Secondly,based on the borrowing records in a university library database,the IBM SPSS Modeler data mining tool was used to cluster readers' borrowing data.The borrowing frequency was used as a parameter to obtain three clustering groups and the frequency of borrowing was explored.High readership.Finally,a correlation analysis algorithm based on matrix data is constructed,which converts string operations based on transactional database into matrix-based Boolean operations,taps strong associations between different books,and places these strong correlation data sets into recommendation sets.,improve recommendation efficiency.The empirical research on the recommendation of library resources in colleges and universities not only can improve the satisfaction of readers,but also can make the university library resources service become passive.The use of data mining results provides specific suggestions for the establishment of a scientific and reasonable library resource structure in university libraries.It is hoped that the utilization of library resources in university libraries will be improved.