An Application Research On University Library System Based On Data Mining

In this information society, huge data accumulated speedily everyday need effective tools to make models and analyze. Data mining is a process to fetch models which is efficiency, novel, latency useful and ultimately understandable. Data mining and data warehouse both are efficiency tools to take knowledge from those very large amount data. In lately years, they are both widely used in every kind trade.Data warehouse' building needs not only all kinds of construction tools but also needs data support. Data warehouse' building is a process of combination all kinds advanced information dealing technologies with enterprise management decisions. If only to combine OLAP, data mining and the huge data in data warehouse with enterprise management decisions, can the data bazaar which costs huge play its greatness function.With the development of the degree of digitalization and with the knowledge and economic times coming, university library is also becoming the center of economic society. It takes on the heavy load of keeping order, spread abroad, development and make use of information/knowledge. Reader's demand is keeping boost and every librarian should be their knowledge navigator, which those challenges press for that university library must find out and master readers' interests and demands, books resources' usage and so on. It should server readers for what they real need, and even can actively induct reader's tendency.Therefore, this dissertation selects university library management system as data warehouse, using data mining to put up data analysis and model making aiming at university library demands. It tries to provide a certain scientific reference during working, management and making decision for workers, supervisor and decision maker.This dissertation introduced the development backgrounds of university library in our country, analyzed the problems which university library being facing with and brought forward the idea of using data mining technology to improve the service level of university library. Firstly we introduced the theoretic knowledge of data warehouse and data mining, based on which we summarized the data and service characteristics of university library. Then we theoretically analyzed the application of data mining on university library from three aspects which included reader, book and service. Secondly we introduced the data warehouse building of university library and gropingly analyzed its important data, out of which we deduced the rules of time using, book sorts and readers' interests for university library. Lastly we made two models; one is using the arithmetic of decision-making tree to set up readers' characteristic model of lower utilizing, the other is using the arithmetic of clustering to build the model of readers' subdivision, from which we educed the characters of different readers. We respect that it can act as a certain reference for future working and forecasting.