Application Research Of Library Management Based On Data Mining Technology

With the rapid development of information industry, most of domestic universities haveadopted the library management system, so that students and teachers through the booksmanagement system management system or network to complete the book’s query andlending. In the query and the borrowing process will accumulate a large number of differenttypes of data for access to library resources, how to use data mining technology to analyzeand process these valuable information, provide high quality service for the readers, providereference for decision makers assisted college libraries need to solve the problem.This paper mainly from two aspects analysis the implementation:On the one hand, to the library to borrow information as the research object, is divided intotwo aspects of readers and books by K-means typical clustering algorithms for data mininganalysis. In the clustering analysis of readers, according to the loan amount of readers, thereader is not active, the general reader, reader active reader of three grades; the day is dividedinto5time periods, readers of each time period, analyses the causes of formation of readerssize; in the clustering analysis on the book in each book, according to the number ofborrowing, the book for readers interested in the book, the general interest books and notinterested in book three. The analysis above all on the library gives reasonable suggestions.The bottleneck problem on the other hand, the Apriori algorithm scans database classicassociation rules algorithm in more times, using a Boolean matrix representation of theoriginal Apriori algorithm based on the improved method. The improved algorithm in thesame thing, the same number of minimum supports time efficiency has been significantlyimproved. And the improved Apriori algorithm to a university library in2013full yearlending records as mining object analysis, find out the strongly related books, and ruleextraction model, the final matching to readers personalized books recommended model thataccording to the rules and the readers.