Machine Learning in Architecture: A Bibliometric Analysis

"Machine learning", which is commonly utilized in engineering, is accepted as a technological innovation having transformative potentials in the architectural discipline as well as in many other disciplines. The possible uses of this technology in architecture have not yet reached a certain level. For this reason, it is extremely important to systematically analyze the related literature, to reveal research trends and gaps, to define new research areas and to reveal possible collaborations that can be established. For this reason, in this study, with a quantitative approach, the researches carried out in the field of architecture through machine learning and the resulting information have been systematically analyzed. For this purpose, the bibliometric data of 461 papers published in the journals scanned through the WoS database in the last 10 years have been used and visualized with a scientific mapping method. With this analysis, the relations between the subjects, the trends in the subject intensity and the collaborations formed by the studies have been revealed through “network maps”. Studies on the use of machine learning in architecture have been evaluated and the possible contributions of a multidisciplinary subject to architectural knowledge have been discussed. As a result, maps in relevant research areas highlighted, possible future research topics shared, and the future impact of this interaction discussed.