Data mining for fraud detection in pawning services offered by banks

Abstract

Pawning or gold-pledged loans has been popular instrument/service among the various products provided by banks and other financial institutions in Sri Lanka in recent decades. Major reason for pawning becoming that much popular is gold items is one of the most reliable sources of credit for lower and middle cast households in Sri Lanka. With the increasing demand for pawning services offered by financial institutions exposure for frauds become incremental accordingly. With the increasing market value of gold people who are seeking for chances to do a fraud are activated and find various methods to achieve their goals. This Paper attempts to explore how data mining techniques can be used in detecting the fraudulent Transactions with related to pawning services offered by banking sector in Sri Lanka.This descriptive analysis on discovering frauds in pawning services will analyses the main parameters associated with pawning and how they can be effectively use for detecting fraudulent transactions. Finally, at the end this research will emphasizes how data mining techniques can be used for detecting fraudulent transactions by providing a solution for pawning fraud detection based on classification model.