Blockchain in Accounting and Auditing: A Systematic Literature Review Using Latent Dirichlet Allocation (An Advanced Machine Learning Approach)

Abstract

The Research is a systematic literature review that reviews 64 peer-reviewed articles published in Google Scholar during the period of 2016 to 2024 to explore the use of Blockchain Technology in Accounting and Auditing. The research outcome relies on Latent Dirichlet Allocation (LDA) topic modeling (A Machine Learning Approach) and computational approach to find out the research publication trends, methodological framework, and thematic areas of focus, geographical aspect in perspective of both trend and descriptive statistics. Results indicate that the highest level of research was reached between 2019 and 2022, and then it started to decrease, which can be explained by the fact that it has already matured to practical application. There is also a high level of focus on data verification and digital validation (51.6% of papers) and a lack of focus on the barriers to adoption and professional views. The overwhelming factor is secondary data (64%), which implies the reference to the exclusion of empirical research. In geographical terms, 80% of the studies assume global views, which exclude jurisdiction-based regulatory, infrastructural and economic conditions essential to the success of implementation. New trends are ERP system integration and blockchain-AI convergence. The gap in research and practice is significant, and not enough focus is given to change management, practitioners' skills, and implementation issues in practice. The future studies should focus on empirical case studies, regional research, and auditor adoption approaches. In addition, there is a lack of research related to barriers or methods of blockchain implementation in the accounting field.