基于客户信用评价模型的营销管理系统

摘要

To improve the quality and efficiency of marketing management for enterprise, A customer credit evaluation model CCEM-MLNs based on Markov logic network is presented. In this model, credit evaluation rules according to the expert knowledge are transformed into the first-order logic rules first so as to compose Markov logic network, then the evaluation model established by training sample data to obtain the weights of all rules. Experimental results show that customer credit classification accuracy rate of this model is increased by 13. 38% in comparison with the model with sharing weights equally. A marketing management system MMS-CCEM based on CCEM-MLNs is designed and implemented. System architecture and hierarchy are given. The system is successfully applied to the marketing management in a Chinese enterprise, and has an obviously effect on customer relationship management.%提高企业营销管理的质量和效率,对客户信用评价模型和企业营销管理系统设计方法进行了研究.提出了一个基于马尔可夫逻辑网的客户信用评价模型CCEM-MLNs,该模型将依据专家知识建立的信用评价规则转化为一阶逻辑规则,构造马尔可夫逻辑网,然后利用样本数据进行训练,得出各规则的权值,建立评价模型.实验结果表明,该模型对客户信用分类的准确率比均分权值的评价模型提高了13.38％；设计并实现了一个基于CCEM-MLNs的营销管理系统MMS-CCEM,给出了系统的体系结构和层次结构,该系统已成功应用于国内某企业的营销管理,在客户关系管理方面效果明显.