ABACUS-A JOURNAL OF ACCOUNTING FINANCE AND BUSINESS STUDIES, cilt.57, sa.3, ss.468-501, 2021 (SSCI)
This study reveals themes and trends in accounting research over the past 20 years by utilizing natural language processing and text-mining techniques. We generated a corpus consisting of over 40,000 articles through multiple searches in EBSCOhost Business Source Premier, Scopus, and ScienceDirect to gather data from 30 highly ranked (A* and A) journals that were listed and categorized by the Australian Business Deans Council (ABDC) as the top accounting journals. Based upon predetermined inclusion and exclusion criteria, we eliminated 24,474 non-empirical articles and those with no abstracts, resulting in 16,449 abstracts. The text-mining analyses reveal 15 distinct clusters, with five clusters showing downward trends, six trending upward, and four maintaining stability. The downward trending clusters are: (1) capital markets; (2) financial reporting; (3) accounting education, careers, and diversity; (4) earnings/markets; and (5) accounting history and capitalism. Trending upward are: (1) critical accounting; (2) auditing; (3) corporate governance; (4) corporate social responsibility; (5) debt financing; and (6) financial markets and forecasting. Stable clusters are: (1) managerial accounting; (2) international accounting standards; (3) taxation; and (4) governmental accounting. This study introduces an innovative method for discerning themes and trends in accounting research and offers a guide to neophyte accounting faculty for determining publishing outlets for research. In utilizing our findings to drill down and provide more detailed knowledge, it also serves as a reference point for future text-mining studies.