Analysis And Application Of The Lending Data On The Universities Library Based On Data-Mining

Library plays an important role, which is the place to store books, periodicals, magazines and audio-visual data, so that teachers and students can not only increase knowledge, seek answers to questions, but also develop reading interests and habits. This paper, through the choice, purification and transfer of lending data provided by the library, probes into the features of library lending data by using data-mining technique, and then puts the research result into the use of library information system. This paper consists of the following aspects.First of all, the paper does some researches on Clustering Analysis Technique and its application in library, has a clustering analysis on the readers and the books using clustering algorithm, produces two lending models and effectively analyzes them in-depth, solves the problem to find books and finds the ways to improve the lending rate. After that, the paper analyzes Association Rules and its application in library, uses improved Apriori algorithm to discover some relations between the books and the readers as well as determine the rationality of books-purchase and further improve books-purchase quality.Then, the paper explores Decision Tree technique supported by clustering and its application in library, uses clustering analysis to obtain the training samples of Decision Tree, and then to obtain high-quality Decision Tree and further improve the preciseness of books'recommendation.Last,taking Anhui Business College of Vocational Technology Library as an example, the paper applies the above research results to analyze lending data. The result of the analysis offers a basis to collection-policy-making, books recommendation and library management for library managers.