Towards a Data Mining Methodology for the Banking Domain

Telecoms and financial service industries are leaders in adopting data analytics technologies, practices, and heavily invest into „Big Data‟ tools and related competence development. However, many of them fail to realize benefits of data-driven decision making and maximize „Big Data‟ business value due to lack of knowledge on how to frame, approach and tackle complex data analytics projects. Existing data mining methodologies are domain-independent, general, abstract and partially outdated. Several refinements of data mining methodologies have been proposed, but they address specific aspects or tasks and remain fragmented. The goal of this doctoral project is to develop a domain-specific data mining methodology for the financial sector, which (1) represents consolidation of existing body of knowledge, and (2) is validated on the sample of real life data-mining projects. The proposed illustrative case studies approach is based on broad, typical data mining use cases portfolio executed across different geographical regions and business areas of the financial institution.