摘要
结合数据立方体技术以及概念分层的分析方法,将面向属性的归纳方法(AOI)与K-means聚类算法相结合,应用于客户时序数据聚类分析中,使每一类客户都具有相似的时序特征。实验表明该方法(AOIGen)能够满足大数据量的客户行为分析要求,比其它方法具有直观、高效等特点。
The data cube technique is associated with concept hierarchy analysis to integrate attribute-oriented induction （AOI） algorithm and K-means clustering algorithm to form an efficient customer clustering algorithm AOIGen. It is applied to clustering analysis of customer time sequence data to make every sort of customers possesses the similar time sequence character. Experimental results indicated that such integration satisfies the demand of customer behaviour analysis with large amount of data, and has the characters of direct perception and high efficiency compared to other algorithms.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第6期126-127,152,共3页
Computer Applications and Software
