Application Research Of Association Rules In Reading Data Processing Of College Library

Along with the rapid development of database technology and the wide application of database management system in the library, massive historical access data to book resources of readers accumulate in the library databases, and massive important information is hidden behind the data. However, the relationships and rules existed in these data cannot be found because of the lack of ability to analyze and integrate data in current library management system. How to mine valuable information from library data using data mining technology has become an important field of application for data mining, and also become a very important research subject in the construction of digital library.In this thesis, in order to validate the feasibility of data mining technology in this field, association rule technology is introduced into the reading data processing system of college library.The main work of the thesis is listed as follows:Firstly, this thesis introduces basic theory and the research status of the data mining, association rules, etc. And two common algorithm for mining association rules-Apriori algorithm and FP algorithm are analyzed. The basic ideas, data mining steps, advantages and disadvantages of both algorithms are discussed, and their performances are evaluated.Secondly, library reader borrowing data is pretreated:including data selection, data integration, data cleaning,data transition and data reduction, and the transaction database for algorithm is established.Finally, the library reader borrowing data is mined using the Microsoft associated algorithm. The rules between readers'characteristic and the book-borrowed and the rues among the types of the book-borrowed are discovered. The experimental results are analyzed and guiding opinions are proposed. Thus some personalized services such as borrow suggestions and book recommendation can be suggested scientifically.