Mapping Two Decades of Artificial Intelligence and MachineLearning in Credit Scoring and Loan Restructuring:A Bibliometric and Network Analysis of Global Research (2000–2025)
Year: 2026
Type: article
Abstract: The past twenty-five years have witnessed a paradigm shift in credit-risk evaluation, driven by the integration of artificial intelligence (AI) and machine learning (ML) into financial decision-making systems. To capture the evolution and intellectual structure of this fast-growing domain, this study conducts a comprehensive bibliometric and network analysis of 87 Scopus-indexed publications published between 2000 and 2025. Using VOSviewer v1.6.20, R-Studio v4.3.1 (bibliometrix 4.1.2), and Python (NetworkX 3.3), the research examines publication growth, geographical distribution, citation impact, co-citation linkages, bibliographic coupling, and keyword co-occurrence patterns. The results show a strong upward publication trend after 2016, reaching its peak in 2024, coinciding with the surge of fintech and explainable AI adoption in banking and credit analytics. The geographical analysis highlights India, China, and the United States as dominant contributors, collectively producing over 60 % of total publications. Citation analysis reveals that a small group of foundational works—particularly Datta (2016) on algorithmic transparency and Ma et al. (2018) on predictive loan modelling—anchor the field’s influence. Author co-citation networks identify three major clusters focused on statistical foundations, credit scoring, and financial distress prediction, while bibliographic coupling uncovers tight cross-linkages among predictive analytics, fintech integration, and explainable ML. Keyword mapping demonstrates a mature conceptual structure with three recurring themes: (i) theoretical development in AI algorithms, (ii) applied ML models in fintech risk management, and (iii) hybrid modelling approaches for credit scoring and interpretability. Together, these findings confirm that the field has evolved from exploratory experimentation to a stage of methodological consolidation and regulatory awareness. The study concludes that future research should emphasize transparency, bias control, and model governance, integrating performance, fairness, and stability as joint evaluation criteria for next-generation credit-risk systems.
DOI: 10.14419/6fv42967
Authors: K. V. Sriranga Abhishek, Dr. G. Sunitha
Institution: National Institute of Technology Warangal
Language: English
Subfield: Accounting
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7124134968