Application of Predictive Analytics in Built Environment Research: A Comprehensive Bibliometric Study to Explore Knowledge Domains and Future Research Agenda

The built environment (BE) sector has seen a significant digital transformation in the past few decades. While predictive analytics (PA) plays a critical role in such a transition. Applications of PA in the BE sector range from making a reliable prediction of the compressive strength of a given concrete mix design, forecasting cost for budget planning, to predicting the safety performance of construction projects. A few literature reviews have been published in related fields but do not provide a comprehensive state-of-the-art review of PA in the BE sector. To fully understand the applications of PA techniques in the BE sector, this study proposes a bibliometric analysis of extant literature. In this paper, a bibliometric analysis of 613 articles in the Scopus database from 2012 to 2021 was carried out to determine the leading knowledge formation entities (countries, institutions, authors, sources, and documents), critical research domains, and the current status of PA-BE research. The results showed that M Y Cheng is the most productive and influential author and China has emerged as the most productive and collaborative country. In terms of scientific impact (measured through average article citation), we observed France to be the most influential country. Through bibliographic coupling analysis, we observed that PA-BE research is presently clustered across four broad verticals: Action Recognition, Sustainability, Cost & Financial prediction, and safety. This paper developed a map of different predictive analytics methods with four clusters and provided PA-BE scholars with research gaps, suggestions and future research directions.