The Application Of Data Mining Technology In Libraries Purchase

The continued progress and development of modern society, knowledge updatesfaster and faster. According to statistics, resources of network information were thegeometric growth in quantity. Web opened to the public increases7.3million pages perday. So fast growth rate requires a modern university library must provide conveniencefor students and teachers who use this information. It is not only for teachers to consultdocuments, but also allow the management department of the books (book of thepurchasing department) to subscribe books validly, to give full play to the library. Thedigital library is a digital information resources system supported by many high-techs,which scatters the interconnection digital information resources in the differentoperators and different parts of the network access method to provide access to theresource for sharing. Data mining and knowledge discovery technology in digital libraryare valuable, which can be divided into three broad categories for different processingobjects: content mining, structure excavation and user log data mining. Almost all of theuniversity libraries have used database management in bibliographic, circulation andreadership data in order to manage the procurement of library books, cataloging andcirculation of books quickly and effectively. The most useful information from thecirculation of books is a large amount of students’ book-borrowing information. Tosome degree borrowing information can be able to stand for the association between thedisciplines and the development of disciplines. Therefore, the aim of this project is todig out the association degree among the various disciplines by using a large number ofreaders borrowing information, to provide help for book purchasing which cancontribute to a higher purpose of the procurement, a higher utilization rate and savingthe expenses of the library. This paper has following tasks:(1) Cleaning, integrating, compressing and conversing the readers’ data to providea clean and effective data source for follow-up work.(2) Using the Apriori algorithm to generate two frequent item sets and pruning, andultimately generate frequent K-itemsets. (3) Using the Apriori algorithm for association mining with the cleaning-out datasource mines out association rules between readers of different grades, occupations andtitles and library resources.(4) Using the Apriori algorithm for association mining with the cleaning-out datasource mines out association rules among some kinds of books of high-borrowing rate.(5) Analyzing the reasons of association rules between the books of computertechnology, English and literature and between them and other kinds of books, andsolving what book to purchase, as well as how to arrange the books.