Current data mining (DM) technology is not domain-specific and therefore rarely generates reliable, business actionable knowledge that can be used to improve the effectiveness of the decision-making process in the banking industry. DM is mainly an autonomous, data-driven process with little focus on domain expertise, constraints, or requirements (Cao & Zhang, The evolution of KDD: Towards domain-driven data mining, 2007). In addition, the predominance of technical metrics rather than domain-specific factors to evaluate the quality of the newly-discovered patterns or relationships limits the chances for discovering useful patterns (Freitas, Are we really discovering "interesting" knowledge from data?, 2006). Consequently, many patterns, rules, or relationships are discovered, but little business-actionable knowledge is generated. Nonetheless, financial institutions continue to rely on DM to gain a competitive edge in the face of rising global competition and eroding profits. Seeking to improve the value of DM for this and other business sectors, experts have suggested shifting from a data-driven to a domain-driven DM approach taking into account domain (business) constraints and requirements and integrating human expertise in all phases of knowledge discovery. Using a quantitative method, the thesis investigated whether and how the integration of human expertise can elicit more useful information from existing DM processes. This study was conducted in two phases. In Phase 1, the relevance of DM research to the banking industry was evaluated using a trend analysis of 6,926 abstracts of articles published over the past sixteen years in business-related DM literature. In Phase 2, the practicality of the banking industry’s current methods for coping with DM limitations was investigated with a particular focus on the role of human expertise in improving the value of DM. For this purpose, a representative sample of 200 data mining specialists working for the banking industry was surveyed. Results, in Phase 1, revealed that current DM research was not relevant to the banking industry. In Phase 2, results showed that the banking industry was ill-prepared in dealing with DM limitations given the less-than-effective use of human expertise to improve the value of current DM technology.