Abstract
Decision Mining (DM) is increasingly gaining attention from academia and slowly progressing towards instrumental application in practice by leveraging decision logs to automatically discover, check for conformance and improve derivation patterns for operational decision-making. This study aims to further operationalize DM by identifying capabilities in the form of functional and non-functional requirements that are posed in the current body of knowledge. By identifying and analysing DM contributions with a focus on derivation patterns we were able to point out the aspects of DM getting attention as well as which did not, e.g., a strong focus on input data and algorithms regarding the discovery phase while the output (data) of the improvement phase seems to be detailed insignificantly. Based on this we formulated a research agenda in which five key points for future research studies are presented.
Publisher
University of Maribor Press