A financial recommender system in digital banking using reinforcement learning and ontology

Given the lack of public trust in financial professional financial advisors due to conflict of interest and the inadequacy of alternative tools in digital banking for financial purpose recommendations, this study designs a framework for an ontological and multifactorial advisory system. To improve financial capability by recommending financial goals, the proposed framework provides an architecture for a personalized financial recommender system designed to identify and recommend specific and achievable financial goals suitable for a wide range of financially savvy users. This framework provides the implementation principles for a digital technology program in digital banking to overcome widespread distrust of traditional financial advisory services as well as instrumental shortcomings in the current outlook for digital banking programs. The system also provides a comprehensive and explicit set of financial target recommendations according to the importance of users. This study empirically examines the usefulness of the proposed framework in terms of effectiveness. The empirical evaluation results show that the program built using this framework is likely to be a reliable and useful program for users to identify and select goals to increase financial capacity.

Keywords:

Digital banking , recommender systems , ontology , machine learning.