It is recommended to choose customer subdivision algorithm according to data character and user intention . lastly an algorithm of measuring interestingness of data mining schema based on user expectation and fuzzy logic is presented 文章对部分应用于crm的数据挖掘算法进行了深入研究，提出根据数据特征和细分目的选择算法，并给出了用于不同标准的客户细分聚类算法。
