Analysis of Correlation Features in Material Application for Architectural Design Based on Big Data Text Mining

Abstract

This study conducted a big data mining and lexical co-occurrence network analysis on design texts sourced from architectural media over the past decade, systematically evaluating the application trends of materials in architectural design. This approach distinguishes itself from traditional case studies that rely on small samples of material applications in architectural design, thereby significantly enhancing the generalizability of the conclusions. The research reveals a 27.2% increase in the material–project frequency of architectural materials over the past ten years, indicating that the materiality of architecture is gradually making a comeback to counteract the trend toward architectural visualization. Within this context, material properties, material interactions, and material presentation in architectural components emerge as universal concerns in architectural design. Notably, the sustainability and energy efficiency of wood and concrete have emerged as pivotal topics in architectural practice, in addition to their structural functions. Additionally, over the past decade, the material–project frequencies of brick and stone have grown rapidly, with respective increases of 24.3% and 11.8%. Brick and stone primarily respond to the locality and style of architectural design by emphasizing their inherent texture and color, thereby enhancing the spatial perception and experience within the designs. This study lays the groundwork for quantitative research on the influence of building materials on architectural design, and it illuminates current trends and preferences in material application within architectural design, thereby empowering architects to make more informed decisions in their material selection.

Keywords

• architectural design texts
• building materials
• big data text mining
• lexical co-occurrence network
• material–design characteristics