This section introduces the method’s structure. First, we will describe the proposed CM-CoP system framework according to the CM-AKD framework.
Next, the proposed predicting customer payment behavior algorithm is described, including using association
rules to produce payment behavior pattern for analysis and derived attributes, utilizing the clustering technique to reduce imbalance, and combing the derived attributes and the attributes provided by the telecom providers to construct the decision treesfor predict user payment behavior patterns.