Data Mining Methodologies in the Banking Domain: A Systematic Literature Review

Abstract

Data mining and advanced analytics methods and techniques usage in research and in business settings have increased exponentially over the last decade. Development and implementation of complex Big Data and advanced analytics projects requires well-defined methodology and processes. However, it remains unclear for what purposes and how data mining methodologies are used in practice and across different industry domains. This paper addresses the need and provides survey in the field of data mining and advanced data analytics methodologies, focusing on their application in the banking domain. By means of systematic literature review we have identified 102 articles and analyzed them in view of addressing three research questions: for what purposes data mining methodologies are used in the banking domain? How are they applied (“as-is” vs adapted)? And what are the goals of adaptations? We have identified that a dominant pattern in the banking industry is to use data mining methodologies “as-is” in order to tackle Customer Relationship Management and Risk Management business problems. However, we have also identified various adaptations of data mining methodologies in the banking domain, and noticed that the number of adaptations is steadily growing. The main adaptation scenarios comprise technology-centric aspects (scalability), business-centric aspects (actionability) and human-centric aspects (mitigating discriminatory effects).