Less effort, more outcomes: Optimising debt recovery with decision trees
- Publication Type:
- Conference Proceeding
- Citation:
- Proceedings - IEEE International Conference on Data Mining, ICDM, 2010, pp. 655 - 660
- Issue Date:
- 2010-12-01
Closed Access
Filename | Description | Size | |||
---|---|---|---|---|---|
2010001669OK.pdf | 1.4 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
This paper presents a real-world application of data mining techniques to optimise debt recovery in social security. The traditional method of contacting a customer for the purpose of putting in place a debt recovery schedule has been an out-bound phone call, and by and large, customers are chosen at random. This obsolete and inefficient method of selecting customers for debt recovery purposes has existed for years and in order to improve this process, decision trees were built to model debt recovery and predict the response of customers if contacted by phone. Test results on historical data show that, the built model is effective to rank customers in their likelihood of entering into a successful debt recovery repayment schedule. If contacting the top 20 per cent of customers in debt, instead of contacting all of them, approximately 50 per cent of repayments would be received. © 2010 IEEE.
Please use this identifier to cite or link to this item: