Sample Research Topic: Credit card fraud detection using big data analytics

Credit card is the dominant method of payment for the last couple of decades. With the wide spread use and ease of use, there are host of security issues which require quick remediation to avoid large financial consequences for the customer and liability for the credit card issuing authority. E-commerce has further spurred the exploitation of credit card by fraudsters.

Babu and Rajeshwari (2016) listed multiple ways to carry out credit frauds namely:

1. ID theft – using name, date of birth and address, attacker can open new accounts

2. Fake cards – skimming machines deployed on ATM or other electronic transaction devices can copy the encoded data from magnetic script and later used in fake card creation

3. Stolen/lost cards – lost wallet with credit cards which are not yet cancelled are easy targets

4. Card not present fraud – some online sites require only credit card numbers & expiry to process transaction which increases the risk of fraudulent use

Babu and Rajeshwari (2016) presented a novel approach using streaming analytics in real time to detect credit card fraud by employing big data analysis tools like Apache Spark, Kafka, HDFS and HBase on Hadoop framework. To detect fraudulent activity, spending behavior, purchase location, date and time of purchase are used and analyzed with historical data to accept or reject the transaction in real time.

Reference:

Rajeshwari, U. & Babu, B. (2016). Real-time credit card fraud detection using streaming analytics. 2nd International Conference on Applied and Theoretical Computing and Communication Technology, 439-444