Mapping the Intellectual Structure of AI-Driven Economic Forecasting: A Scientometric Analysis From 1991 to 2024

This study presents a comprehensive bibliometric analysis of artificial intelligence (AI) applications in economic forecasting from 1991 to 2024. Through systematic examination of 746 publications from the Web of Science Core Collection, we employ collaboration network analysis, co-citation analysis, and keyword co-occurrence analysis to map the intellectual structure and development trajectory of this rapidly growing field. Our findings reveal distinct patterns in international research collaboration, with China and the United States emerging as primary contributors, while European institutions demonstrate strong centrality in global research networks. The co-citation analysis identifies five major research clusters, highlighting the field’s theoretical foundations in explainable AI, deep learning applications, support vector regression, specialized forecasting domains, and behavioral finance integration. Temporal analysis of keyword co-occurrence patterns indicates an evolution from basic neural network applications to sophisticated hybrid approaches incorporating multiple AI techniques. The study provides novel insights into emerging research frontiers, particularly in areas of explainable AI, privacy-preserving computation, and adaptive modeling for economic forecasting. This comprehensive analysis contributes to the literature by mapping the field’s intellectual landscape and highlighting promising future research directions, providing valuable guidance for researchers, practitioners, and policymakers working at the intersection of AI and economic forecasting.