A topical network based analysis and visualization of global research trends on green building from 1990 to 2020

Abstract

The explosive growth of green building (GB) research over the past few decades has posed great challenges for researchers to effectively grasp the holistic GB research status. This paper presents a text mining based method to identify the key research topics and trends from the existing abundant GB research publications. In total, 2421 articles are retrieved from the Web of Science core database. Subsequently, three mutually integrated text mining techniques, including the latent Dirichlet allocation (LDA) modeling, Word2vec, and community network analysis, are employed for knowledge discovery. Results show that the number of GB research topics increased from 1990 to 2020, while design optimization and energy efficient measures have always been the research hotspots. The identified research themes are further categorized into nine general research areas, including GB design, energy saving, GB rating system, life cycle evaluation, incentive and hindrances, post-occupancy evaluation, GB technologies, GB market, and management aspects of GB. Based on the revealed emerging research themes in recent ten years, three research directions are further proposed, such as green building market, employment of digital technologies, and safety risks in GB projects. This research provides an applicable quantitative method to efficiently identify the research topics from the extensive research publications and establishes a comprehensive knowledge framework of the GB research status at different stages.