Font Size: a A A

Based On Data Mining Library Personalized Recommendation System

Posted on:2014-10-02Degree:MasterType:Thesis
Country:ChinaCandidate:F X NieFull Text:PDF
GTID:2268330401454469Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of network and digital libraries, the digital age of university libraries is coming.At present, there are massive amount of date stored in library automation management system, which contains bibliographic data, the reader circulation data, bibliographic records and Web access records.But the university library information management system with simple statistical analysis function cannot predict the demands of readers and tendency of books to borrow. Data mining techniques can be useful knowledge to find hidden information from this massive data source. In this paper, association rules of data mining technology and clustering analysis technique has been systematically studied and we apply the research results to the university library data mining work so that we can analyze Readers’ interest and the utilization of the books and finally design a personalized recommendation system for the Northwestern University Library.Based on the research of date mining algorithm, we amylase and forecast the date of the Northwestern University Library automation management system in this paper. The progress of the main work is as follows:(1)Analyzing the basic statement of the date mining technology’s application in library and giving detail summary about the functions used in the Northwestern University Library.(2)Applying association rules and clustering analysis to the Northwestern University library automation management system and mining massive date.(3)The system has two functions:One is reader recommended function, according to the excavation of the reader circulation records and situation of similar books borrowing, the books which readers are likely interested in should be recommended; The other is administrator suggestion function, according to the excavation of the circulation date, web query data and bibliographic retrieve date, the system can recommend the books with high rate of borrowing or circulation to the administrator so that book purchasing will be done by the scientific and efficient way.
Keywords/Search Tags:data mining, association rules, clustering analysis, library, recommendationsystem
PDF Full Text Request
Related items