A genetic-based data mining model for customer relationship management in B2B ecommerce
Customer Relationship Management (CRM) improves the responsiveness and understanding of business strategies among employees and helps to achieve better customer service. However, in Business-to-Business (B2B) eCommerce, CRM data are rarely analyzed across market segments or customer categories, and customer-to-business relationship also poses issues. Thus, making appropriate decisions in the CRM model is difficult. This study presents a modified B2B CRM using the Genetic algorithm and Data Mining Techniques to improve decision making. The model classifies consumers into consumers of Repeat and Shop-and-Go. Modified data mining C5.0 and the Genetic algorithm was employed to optimize rules generated by the decision tree algorithm. The findings showed that the proposed model allocates resources effectively to the most profitable customers’ decisions. The output metrics are machine time, calibration graph, and ROC curve. In comparison with the conventional C5.0, k-NN, and Support Vector Machine, the proposed model has greater accuracy of 89.3 percent.
Keywords: Customer Relationship Management, B2B eCommerce, Genetic Algorithm, Data Mining