臺灣博碩士論文加值系統

報表通常被使用來調整催收系統的催收策略並做為日常催收作業的補充資訊。雖然這是用來找尋有效催收策略的常用方法，但使用報表來輔助日常催收作業可能會抵消使用催收系統所帶來的好處。為了能真正改善催收系統的效能，本研究提出了經修正的催收系統架構。該架構運用條件成熟資訊及轉換權值來揭露每一催收案件可採用的催收行動的有效程度。導入該架構後的催收系統，不只能導引催收人員在每一案件上採取最有效的行動，更可有效地消除人為因素及精進催收策略，從而提昇整體的催收績效。

The reports are usually used to refine the collection strategies and also as the supplementary information of the daily operations of the generic debt collection. Although it is a common approach of finding the effective and efficient collection strategies by using the reports, it is a bad idea to use the reports to assist the daily operations because it probably cancels out the advantage of using debt collection system. In order to improve the performance of the generic debt collection system, this research proposes a modified architecture of the generic debt collection system. The proposed architecture uses the conditional mutual information and transformed weight to discover the effectiveness of every action of every case. By adapting the proposed architecture, the debt collection systems can guide the collectors to do the most effective action on every step of every case. Finally, it improves the overall performance of debt collection system obviously not only by eliminating the personal equations but also by refining the collection strategies.

Table of Contents
口試委員會審定書 ii
授權書 iii
Abstract (in Chinese) iv
Abstract (in English) v
Acknowledgement vi
List of Figures vii
List of Tables viii
Table of Contents ix
Chapter 1 Introduction 1
1.1 BACKGROUND 1
1.2 OBJECTIVE 2
1.3 ORGANIZATION OF THESIS 4
Chapter 2 Operation of Debt Collection Business 5
2.1 COLLECTING PROCEDURE 5
2.2 DECISION-MAKING 6
2.3 GENERIC DEBT COLLECTION SYSTEM 7
Chapter 3 Modified Architecture of Debt Collection System 10
3.1 DEFICIENCIES OF GENERIC DEBT COLLECTION SYSTEM 10
3.2 MODIFIED ARCHITECTURE 10
3.3 IMPORTANCE OF PERFORMANCE STATISTICS 12
3.4 APPROACH OF CREATING PERFORMANCE STATISTICS 13
3.4.1 GET ACTION AND RESULT TOGETHER 15
3.4.2 COMPUTATION 16
3.4.3 WEIGHTING 21
3.5 FILTERING AND RANKING 21
Chapter 4 Pilot Scheme 25
4.1 OVERVIEW 25
4.2 EXPERIMENTAL ENVIRONMENT 25
4.3 EXPERIMENT OF CREATING PERFORMANCE STATISTICS 26
4.4 EXPERIMENT OF RANKING MECHANISM FOR ALL CASES 28
4.5 EXPERIMENT OF FILTERING AND RANKING MECHANISM FOR EVERY CASE 29
4.6 DISSCUSSION 31
Chapter 5 Conclusion 32
5.1 SUMMARY 32
5.2 RELATED WORKS 33
Reference 34

REFERENCES
[1] Non Performing Loan ratios of Domestic Banks, announced by Central Bank of China. http://www.cbc.gov.tw/bankexam/data/ratios.doc
[2] CACS®, developed by AMS Corporate, USA. http://www.amsinc.com/creditrisk/products.asp#CACS
[3] DebtManager, developed by London Bridge Group, UK. http://ea.london-bridge.com/products/index.cfm?Product_id=11
[4] Mark Last and Oded Maimon, “A Compact and Accurate Model for Classification”,
IEEE Transactions on Knowledge and Data Engineering, Vol. 16, No. 2, pp.203 - 215, Feb. 2004
[5] Efraim Turban and Jay E. Aronson, “Decision Support Systems and Intelligent Systems” Sixth Edition, Prentice Hall, USA.
[6] David S. Moore, “Statistics: Concepts and Controversies”, Fifth Edition,
W. H. Freeman & Company, Inc., USA.
[7] Jody J. Daniels and Edwina L. Rissland, “Integrating IR and CBR to Locate Relevant Texts and Passages”, Proceedings of the International Workshop on Database and Expert System Applications (DEXA ’97)
[8] InduShobha N. Chengalur-Smith, Donald P. Ballou, and Harold L. Pazer,
“The Impact of Data Quality Information on Decision Making: An Exploratory Analysis”, IEEE Transactions on Knowledge and Data Engineering, Vol. 11, No. 6, pp. 853 - 864, Nov.-Dec. 1999.
[9] Stephen T. C. Wong, “Preference-Based Decision Making for Cooperative Knowledge-Based Systems”, ACM Transactions on Information Systems, Vol. 12, No. 4, pp. 407-435, Oct. 1994.
[10] João W. Cangussu, Raymond A. DeCarlo, and Aditya P. Mathur, “A Formal Model of the Software Test Process”, IEEE Transactions on Software Engineering, Vol. 28, No. 8,
pp. 782-796, Aug. 2002.