摘要
针对传统的流通数据关联规则检测方法存在检测时间较长、准确率较低等问题,提出一种基于Kullback-Leibler散度的定期借阅图书流通数据关联规则检测方法。将定期借阅图书流通数据符号化,并利用符号序列的频繁序列作为图书流通数据关联度的衡量指标,计算图书流通数据频繁序列中两个特征项之间的相对熵,将其作为图书流通数据频繁序列特征项之间流通数据近似度。利用平均相对熵模型计算频繁序列中全部特征项与频繁序列主题之间的关联度;对得到的关联度值采取加权求和操作,获得两个图书流通数据之间的关联规则,实现定期借阅图书流通数据关联规则检测。实验结果表明,所提方法能够快速、准确地对图书流通数据进行关联规则检测。
Traditionally, the detection method leads to long detection time and low accuracy. Therefore, a method to detect the association rule of regular borrowing book circulation data based on Kullback-Leibler divergence was proposed. Firstly, this research symbolized the circulation data of regular borrowing book and used the frequent sequence of symbol sequence as the measure of association degree of book circulation data. Secondly, our research calculated the relative entropy between two feature items in the frequent sequence of book circulation data, which was used as the approximate degree of circulation data between sequence feature items of book circulation data. Then, the research used average relative entropy model to calculate the association degree between all feature items in frequent sequence and the frequent sequence topics. Moreover, the weighted summation operation was performed on the association degree to obtain the association rule between two book circulation data. Finally, the association rule detection for regular borrowing book circulation data was achieved. Simulation results show that the proposed method can quickly and accurately detect the association rule of book circulation data.
作者
李萍
汪滢
LI Ping;WANG Ying(College of Science and Technology,Nanchang University,Nanchang Jiangxi 330029,China)
出处
《计算机仿真》
北大核心
2020年第5期354-357,452,共5页
Computer Simulation
基金
江西省教育厅科学技术研究项目(GJJ171458)。
关键词
定期借阅图书
流通数据
关联规则检测
Regular borrowing of books
Circulation data
Association rule detection