The Evolution of Possibilistic Risk Research: A Bibliometric and Topic-Modeling Perspective

Abstract This study aims to provide a comprehensive and data-driven analysis of the evolution, intellectual structure, and thematic dynamics of possibilistic risk research over the period 1975–2025. A dataset of 461 documents retrieved from the Web of Science Core Collection was analyzed using bibliometric science-mapping techniques implemented in the Bibliometrix R package. To complement the structural analysis, Latent Dirichlet Allocation (LDA) and Structural Topic Modeling (STM) were applied to titles and abstracts. Topic selection was based on a combination of perplexity, UMass coherence, and qualitative interpretability of topics, while STM incorporated publication year to capture temporal dynamics. The results reveal a steady growth of scientific production and an increasingly internationalized research landscape. Bibliometric and network analyses highlight a well-defined intellectual core rooted in possibility theory and fuzzy uncertainty, alongside strong expansion into applied domains. Topic modeling identifies eight major themes, showing a clear shift from foundational theoretical work toward optimization, computational methods, and application-oriented research, particularly in supply chains, decision-making, and complex systems. Temporal analysis confirms the declining dominance of purely theoretical topics and the rise of interdisciplinary and application-driven research. The findings demonstrate that possibilistic risk research has evolved into a mature and methodologically integrated field, combining strong theoretical foundations with growing practical relevance. By integrating bibliometric and topic-modeling approaches, this study provides a unified and robust perspective on the field’s development and underscores the increasing role of optimization, data-driven methods, and interdisciplinary applications in shaping future research directions.