Data Mining Approach In Financial Fraud Detection and a Literature Review

Abstract

Only in USA Stock Exchanges, daily avarage trading volume is about 7 billion units. Just depending on this statistics, the necessity of information discovery hidden in data is a reality to tackle the problems in strategic, tactical and operational decision processes with lower costs and higher reliability. Information discovery from databases, namely, data mining is an effective method for auditing financial statements in companies and providing flow of information between market players as well as risk and portfolio management as banking applications. In this study, 79 journal articles related to the subject published 1994-2015 have been classified and evaluated in terms of data mining techniques. It has been found that data mining techniques have been applied most extensively to detection of banking and insurance fraud. However, the findings of literature review show that the number of studies in detection and prediction of financial fraud is not enough in Turkey.