Machine Learning in Management Accounting Research : Literature Review and Pathways for the Future

Abstract

This paper explores the possibilities of employing machine learning (ML) methods and new data sources in management accounting (MA) research. A review of current accounting and related research reveals that ML methods in MA are still in their infancy. However, a review of recently published ML research from related fields reveals several new opportunities to utilize ML in MA research. We suggest that the most promising areas to employ ML methods in MA research lie in (1) the exploitation of the rich potential of various textual data sources; (2) the quantification of qualitative and unstructured data to create new measures; (3) the creation of better estimates and predictions; and (4) the use of explainable AI to interpret ML models in detail. ML methods can play a crucial role in MA research by creating, developing, and refining theories through induction and abduction, as well as by providing tools for interventionist studies.