jaue2021-004: Machine Learning Applications in Architecture: A Bibliometric Analysis

Abstract

As the core research of artificial intelligence and data science, machine learning (ML) has been emerged as an effective way to achieve artificial intelligence applied in the field of architecture. However, it is still not clear how it can effectively help architects to solve which problems in the reality. In this paper, we first analyze the knowledge mapping of relevant research in the CumInCAD platform and clarify the related research clusters. Based on this, the distribution of research hotspots and regions are then analyzed. Secondly, a more comprehensive cross-sectional analysis of research tasks and relevant algorithms which have been applied in recent years is presented. Based on neural network, clustering, dimension reduction, regression, classification and other relevant algorithms, the paper discussed their applications in different research tasks. Finally, the effectiveness and trends of the application of neural network algorithms are evaluated, to present a more objective trajectory of ML in the field of architecture, while providing a more comprehensive research basis for future research and researchers.