Discovering Architectural Dynamic Knowledge: An Extension Clustering Approach
Qiang Guo, Guangtian Zou, Pan Liao · 2018
Architectural technology has developed rapidly in a wide spectrum of fields in recent years, from theories, design strategies, spatial forms to materials. In order to gain an in-depth understanding of the trend in the cutting-edge technology, a method of discovering Architectural Dynamic Knowledge (ADK) based on Extension Clustering (EC) was proposed in this paper. To achieve this, firstly, web data representing the trend in the architectural domain were collected through data acquisition software. The data were then processed so as to build an Architectural Dynamic Knowledge Database (ADKD). Secondly, extensible analysis methods and reversed thinking modes were conducted to analyze the innovation problems in current architectural projects, providing theory support for the ADK discovery process. Finally, high-frequency words associated with formed clusters were obtained using the EC methods. Analyses of the relationships between clusters and words in each cluster were undertaken to further discover ADK in the architectural domain. The case studied in this paper confirms that the proposed method can increase architects' abilities in terms of utilizing web data and discovering ADK, thus accelerating the process of computer-aided architectural design innovation.