Akademik Veri Yönetim Sistemi
On Architecture - Shaping The City Through Architecture, Belgrade, Sırbistan, 5 - 06 Aralık 2024, ss.45-53, (Tam Metin Bildiri)
Today, concerns about climate change, environmental problems, and resource depletion have become a significant problem on a global scale, especially with increasing energy consumption and the role of built environments in this context. Buildings, which constitute a large part of the built environment, have been recognized as the main responsible due to their increasing energy needs and the emissions resulting from their energy consumption. For this reason, within the scope of sustainable architecture, which has occupied the agenda as a solution approach, energy conservation, which aims to ensure energy efficiency in buildings, reduction of carbon footprint, and the potential for the use of renewable energy sources have started to gain importance in architectural research. However, achieving energy efficiency and meeting other sustainability goals is a complex process that requires simultaneous evaluation of many parameters such as technical, aesthetic, economic, etc., and multiple decision-making. In this context, by integrating today's computer and software technologies into various fields, the potential of machine learning, which can learn by associating various data and make optimal predictions based on the information it has learned, stands out in the discipline of architecture. Machine learning, which has found a place in architectural research, especially in building energy system modeling and analysis, offers powerful algorithms and computational methods for processing and analyzing complex data sets and optimizing multiple variables. In this study, which aims to examine academic research on the use of machine learning in the context of sustainability and energy efficiency in the field of architecture, a bibliometric analysis was conducted using keywords related to the subject through the Web of Science database. Data such as the annual number of papers, the relationships between authors, and the relationships between documents were analyzed, and the frequency and distribution of terms and concepts within the scope of the subject were examined. A cluster map of the analysis results was created using VOSviewer and RStudio software. The study identified the topics on which the machine learning methods used in the papers were focused, in the context of sustainable and energy-efficient building parameters, and highlighted the potential applications of machine learning in these areas. This study can serve as a reference for future research on using machine learning to improve building energy efficiency.