Based on MMS_Cumulate algorithm and GP-Apriori algorithm, a data mining algorithm, MAR_LCR is proposed for library new book recommendation service which is capable of finding generalized association rules in the form of “reader-book” and allows the user to specify multiple minimum supports for different items. The search space is greatly cut down by improving the process of candidate generation. Experiment results show that the MAR_LCR algorithm is highly effective. Finally, a new book recommendation model is proposed.
佘俊胜,黄战 . 一种面向图书馆新书推荐服务的广义关联规则挖掘算法[J]. 现代图书情报技术, 2006, 1(10): 16-20.
She Junsheng,Huang Zhan . A Generalized Association Rule Mining Algorithm for Library New Book Recommendation. New Technology of Library and Information Service, 2006, 1(10): 16-20.
1贾春华. 数字图书馆的个性化信息服务. 现代图书情报技术， 2004(12):88-90
2R. Srikant, R. Agrawal, Mining generalized association rules, Proc. of the 21st VLDB Conf., 1995.409-419
3B. Liu, W. Hsu, Y. Ma Mining association rules with multiple minimum supports,Proc. ACM SIGKDD Conf. on KDD, San Diego, 1999.337-341
4Ming-Cheng Tseng, Wen-Yang Lin. Mining Generalized Association Rules with Multiple Minimum Supports. DaWaK, 2001. 11-20
5S.Y. Hwang, E.P. Lim, A data mining approach to new library book recommendations, Lecture Notes in Computer Science, Singapore, 2002.229-240
6S.-Y. Hwang, C-W. Tsai, Targeted Advertising Based on GP-association Rules: ［Dissertation］. Guangzhou:Sun Yat-Sen University, 2001
7司徒浩臻. 数据挖掘技术在图书馆信息服务中的应用. 现代图书情报技术, 2005(10):15-18
8中国图书馆图书分类法编辑委员会. 中国图书资料分类法. 北京：科学技术文献出版社, 2000
