Integrating of business intelligence and CRM in banks: An empirical study of SOM applied in personal customer loans in Taiwan

According to the development of economy, the requirements of personal consumer loans have increased. However, the banks are often worried about the status of repayment because the behaviors of every borrower are always fuzzy and uncertain. The most important thing is banks must realize the credit risk of borrowers when they review the applications of personal consumer loans. As a result, this study aims to mine the patterns of default in big data of the banks. The method of self-organizing mapping (SOM) was applied to divide the borrowers into different groups. The result showed we found 5 groups of borrowers and the 5 degrees of default. Finally, this study analyzed the characteristics of borrowers in each groups, and provides the appropriate strategy to individual groups to lower the credit risk in personal consumer loans.