Abstract: Simulating the market demand and supply interaction on individual banks and applying econometric techniques on the simulated bank data to statistically establish a significant operational relationship between the banking industry and the real sector of the economy. Stress-testing models have also been developed for some aggregate banking sector indicators for work. Cost and benefits of fraud detection solutions were evaluated to substantiate the business justification and return on investment for building, maintaining and supporting fraud detection solutions. The NIST Big Data Interoperability Framework and the DAMA-DMBOK framework were utilized to define the data ecosystem along six perspectives—data sources, data governance, data quality, data storage, data privacy and data provenance. The analysis included a taxonomy of financial fraud, and the definitions and characteristics of fraud detection, risk assessment and fraud monitoring. Some of the trends emerging in the digital banking ecosystem were also discussed. Banks need to achieve a balance in the prevention and detection of fraud, which is seen as never ending and deserves constant attention. The playbook defines how the operation teams respond to fraud. Such resources should ideally reside in the same location and have clear visibility for decision-making.

Given the extensive amount of available data, fraud detection solutions for banks require solid scientific foundations. Data-driven approaches that leverage machine learning and data-mining techniques are therefore being explored as tools for improving the prediction and detection of fraud events. Supervisory authorities require banks to incorporate fraud detection systems into their digital banking ecosystems. Such Data Driven Decision Systems must be capable of automatically detecting suspicious activity within acceptable limits of cost, accuracy, risk, coverage and performance level without too many false positives and as few wrong flags as possible.

Keywords: Banking; Big Data; Cloud Computing; Data Mining; Digital Forensics; Decision Theory; Risk Management; Real-time Analytics; Telecommunication; Fraud Detection.eva.

Downloads: | DOI: 10.17148/IJIREEICE.2020.81213