Data-driven literature analysis of artificial intelligence (AI) in real estate research: a bibliometric and machine learning-based approach

Purpose This paper aims to provide a comprehensive review of research trends and developments in the application of artificial intelligence (AI) within the real estate domain, with a focus on understanding the field's diversity and evolution. Design/methodology/approach A total of 600 publications related to AI in real estate, published between 1988 and 2024, were retrieved from Scopus. This research conducted bibliometric analyses using SCImago Graphica, Bibliometrix and ScientoPy to examine global publication patterns and research trends. To identify underlying topics and guide future research, Structural Topic Modelling (STM), a machine learning (ML)-based approach, was applied to uncover latent themes. Findings The analysis reveals a growing interest in AI applications in real estate, especially following the recent AI boom. The United States and China lead in research output, each contributing over 80 publications since 1988. Results from the ML analysis highlight several emerging research themes, including “Deep Learning Models for Property Analysis”, “Virtual/Augmented Reality Applications”, “PropTech and Smart/Digital Systems in Property Practices” and “Smart Urban Planning and Land Use”, all of which present considerable potential for future research. Originality/value This study provides a comprehensive overview of the literature on AI in real estate. By applying STM, it introduces a novel approach to reviewing large volumes of literature. STM enables the synthesis of complex information and reveals key research gaps and emerging thematic trends. The insights gained can inform strategic decision-making for both academia and industry professionals.