改进的FP-growth关联规则算法及其在图书推荐系统中的应用

The Improve and Application of FP-Growth Algorithm of Association Rules

在FP-growth关联规则算法的基础上提出了基于动态二维数组的算法,引入可变二维数组结构,动态的将事务数据库存入该数组中,可以大大提高数据挖掘的效率。并以图书馆管理系统中的图书借阅数据作为训练数据,使用改进的FP-growth算法实现了高校图书推荐系统,本系统能够从图书馆图书借阅记录中挖掘和发现读者借阅行为中隐含的规律,得到读者与图书的频繁项集,从而可以实现对不同身份的读者推荐不同类型的图书功能。

Based on the algorithm of FP-growth, the paper proposed a dynamic two-dimensional array. The algorithm leads in the variable two-dimensional array structure and stores the dramatic transaction database into the array. It significantly improves the efficiency of data mining, meanwhile, by using library’s managing systems’ book lending data as a training data. The paper uses the improved FP - growth algorithm and has accomplished to the book recommendation system in colleges and universities. From the lending record in the library, this system can explore and find the rules from readers’ behavior as well as frequent itemsets between readers and books, thus it manages to recommend different types of books to readers of different identity.