Introduction
The growth of eCommerce systems has led to an increase in online transactions using credit cards and other methods of payment services. These transactions mainly require that a user provide their personally identifiable information, which significantly increases risks associated with credit card frauds and identity theft. The article by Chen et al. (2015) introduced the concept of big data fraud and the strategies employed by Alibaba to curb cases of online fraud. The study explored different mechanisms employed in fraud risk monitoring and management systems, such as statistical and engineering techniques. More specifically, Chen et al. (2015) study described the Counter-Terrorism Unit (CTU), AntBuckler, and RAIN score models used to monitor, identify, and block online fraudsters.
Possibility of Implementing Alibaba Style of Fraud Management in U.S. based Institutions
Cases of online fraud affect all the people who use the Internet worldwide. A study by Mahmoudi and Duman (2015) revealed that in the United States, 70% of people who perform online transactions have security and identity fraud concerns. Additionally, Mahmoudi and Duman (2015) reported that, in 2012, cases of online fraud in the U.S. led to a loss of $3.5 billion. Therefore, the US-based institutions would benefit from Alibaba’s style of fraud management.
Online activities usually leave a digital footprint of their activities. Fraud detection systems use various algorithms to monitor and classify these activities as normal or malicious. For instance, the CTU system described in Chen et al. (2015) emphasizes the assessment of historical cases and user behaviors to identify patterns that would classify an online user as a fraudster. Examples of techniques used in fraud detection include the Naïve Bayes method and the K-nearest neighbor (KNN) technique (Zareapoor & Shamsolmoali, 2015). The fraud management systems monitor the activities of online users to identify patterns, which are then used to predict the possibility of fraudsters. The systems would then report or block the user from accessing online systems, subsequently saving companies from experiencing losses associated with online fraud.
Aspects of Framework to be impacted
Alibaba’s fraud risk system has five layers essential in the identification and prevention of attacks. The five layers as described in the Chen et al. (2015) study are account check, deice check, activity check, risk strategy, and manual review. Therefore, implementation of the system in the U.S. based company will result in changes as to how customers and transactions are processed. In the account check, the details of the customers will be checked against those existing in database to identify if a customer is listed as fraudulent. If the customer passes this layer, then the details of the device (computer/phone/tablet) are analyzed to assess their online activities. Devices with a bad history are blocked. The third layer assesses the users’ online behaviors, including linked accounts and device history. Users with historical patterns or behaviors that suggest fraud are also blocked. The fourth layer is concerned with judgment of these activities and making necessary actions (Chen et al., 2015). Depending on the severity of the violations, users are blocked, or cases are sent to manual review where the customer service manually review the customers’ activities.
References
Chen, J., Tao, Y., Wang, H., & Chen, T. (2015). Big data based fraud risk management at Alibaba. The Journal of Finance and Data Science, 1(1), 1-10. Web.
Mahmoudi, N., & Duman, E. (2015). Detecting credit card fraud by modified Fisher discriminant analysis. Expert Systems with Applications, 42(5), 2510-2516. Web.
Zareapoor, M., & Shamsolmoali, P. (2015). Application of credit card fraud detection: Based on bagging ensemble classifier. Procedia Computer Science, 48, 679-685. Web.