Comparing research frontiers: A machine-learning analysis of the research published in the Accounting & Finance Review and in international leading accounting journals
Open Access
- 1 January 2026
- journal article
- Published by FapUNIFESP (SciELO) in Revista Contabilidade & Finanças
- Vol. 37 (100)
Abstract
This study compares the studies published in the Accounting & Finance Review (A&FR) with those of top-tier international accounting journals over the past decade. It aims to offer a comprehensive overview of the topics under analysis in each of these settings, highlighting key areas of research activity and ongoing debates in the accounting field across both outlets. By mapping and comparing the thematic landscape of the A&FR with that of top-tier international journals, the study provides a structured understanding of how accounting research is developed across different publication outlets. These results could potentially serve as a foundation for setting future research agendas and guiding academic inquiry toward areas of high relevance and potential impact. Our text-mining approach enables the discovery of patterns and emerging trends that are often overlooked in manual reviews. The insights from this analysis can inform academic strategy, support curriculum development, and guide funding priorities, ultimately fostering more targeted and impactful research within the accounting discipline. Our analysis is based on a dataset of 277 articles published in the A&FR and 3,652 articles published in top-tier international accounting journals classified as Association of Business Schools (ABS) 4* and ABS 4 from 2015 to 2025. Using Python, we apply the latent Dirichlet allocation (LDA) algorithm to extract topics from the abstracts, identifying latent patterns and thematic structures within the research corpus. For the A&FR, the model identifies nine distinct topics: Control systems, Budgeting-related issues, Organizational performance, Funding and credit-related matters, Insurance and actuarial issues, Taxation, Earnings management, Auditing, and Financial markets and investment returns. In top-tier international journals, the LDA solution identifies seven topics: Auditing, Earnings announcements and analyst earnings forecasts, Corporate disclosure, Accounting and capital markets, Taxation, Management control systems, and Empirical modelling in financial accounting. Despite differences in emphasis, both outlets share two topics (Auditing and Taxation), indicating at least some areas of convergence between Brazilian and international accounting scholarship.Keywords
This publication has 139 references indexed in Scilit:
- Automatic detection of relevant information, predictions and forecasts in financial news through topic modelling with Latent Dirichlet AllocationApplied Intelligence, 2023
- Latent Dirichlet allocation (LDA) and topic modeling: models, applications, a surveyMultimedia Tools and Applications, 2018
- Consequences of adopting an expanded auditor’s report in the United KingdomReview of Accounting Studies, 2018
- The taxman cometh: Does tax uncertainty affect corporate cash holdings?Review of Accounting Studies, 2017
- Seeing is believing: analysts’ corporate site visitsReview of Accounting Studies, 2016
- What do accruals tell us about future cash flows?Review of Accounting Studies, 2016
- Accruals and price crashesReview of Accounting Studies, 2016
- Speaking of the short-term: disclosure horizon and managerial myopiaReview of Accounting Studies, 2015
- A Text Mining-Based Review of Cause-Related Marketing LiteratureJournal of Business Ethics, 2015
- The effect of target-firm accounting quality on valuation in acquisitionsReview of Accounting Studies, 2014