Feasibility Analysis of Data Science Methodologies in Architectural Design

The integration of data science into architectural design has revolutionized the creation of innovative, sustainable, and efficient buildings; however, selecting a practical methodology for such projects remains underexplored. This study conducts a thorough comparison of prominent data science methodologies-CRISP-DM, Agile Data Science, Microsoft's Team Data Science Process (TDSP), and OSEMN-to identify the most appropriate framework for architectural design. These methodologies are evaluated based on criteria tailored to architecture's unique needs, including alignment with business understanding, data preparation capability, modeling and simulation, support for team collaboration, flexibility, and deployment and monitoring. The analysis reveals that TDSP excels due to its balanced approach, combining structured planning with flexibility, making it ideal for the complexities of architectural workflows. TDSP supports advanced modeling for energy optimization and structural analysis while ensuring effective deployment and monitoring in real-world scenarios. In contrast, CRISP-DM offers a robust structure but lacks adaptability, Agile Data Science prioritizes flexibility over depth, and OSEMN proves too simplistic for large-scale projects. These findings provide actionable insights for architects and data scientists seeking to leverage data-driven techniques while also identifying opportunities for future research into integrating emerging technologies with TDSP. Ultimately, this study highlights the transformative potential of data science in architectural design, with TDSP emerging as a standout methodology for optimizing project outcomes.

کلیدواژه ها:

Data Science ، Architectural Design ، Methodology Comparison ، Team Data Science Process (TDSP)