Personalized requirements oriented data mining and implementation for college libraries
Zhao Xiang1, Zheng Hao2
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 293-300
1Shandong University of Finance and Economics Library, Shandong Jinan 250014, China;
2School of Business Administration, Shandong University of Finance and Economics, Shandong Jinan 250014
By the study of data mining technologies and systematic theory of personalized service, this paper introduces data mining to the college libraries to provide personalized requirements oriented service for readers. It analyses the demand for college library database at the phase of data mining, and explains the necessary for modelling in theory. Then the structure of model is designed. During data mining, we adopt ClassIndex number to establish index distribution trees. We compute the interest distance among the readers according to the depth of ClassIndex numbers of books. Inspired by Cruskal’s method, we use minimum spanning tree to establish a weighted undirected liaison graph to perform clustering analysis for the readers. In the association rule mining, by the clustering of readers’ borrowing information we find the results are ideal. So we can offer corresponding rules pattern to provide personalized recommendation service for readers.