Mixed-initiative synthesized learning approach for web-based CRM

The issue of customer relationship management has emerged rapidly. Customer s have become one of the most important considerations to new companies bei ng built. Accordingly, customer retention is a very important topic. In thi s paper, we present a mixed-initiative synthesized learning approach for be tter understanding of customers and the provision of clues for improving cu stomer relationships based on different sources of web customer data. The a pproach is a combination of hierarchical automatic labeling SOM, decision t ree, cross-class analysis, and human tacit experience. The objective of thi s approach is to hierarchically segment data sources into clusters, automat ically label the features of the clusters, discover the characteristics of normal, defected and possibly defected clusters of customers, and provide c lues for gaining customer retention. (C) 2001 Elsevier Science Ltd. All rig hts reserved.