The STG Pattern: Application of a “Semantic-Topological-Geometric” Information Conversion Pattern to Knowledge Modeling in Architectural Conceptual Design

ABSTRACTGenerative modeling tools have become a popular means of composing algorithms to generate complex building forms at the conceptual design stage. However, composing algorithms in order to meet the requirements of general design criteria, and communicating those criteria with other disciplines by means of generative algorithms still faces many technical challenges. This paper proposes the use of a “Semantic-Topological-Geometric (STG)” pattern to guide architects in composing algorithms for representing, modeling, and validating design knowledge and criteria. The STG pattern can help architects to convert semantic information concerning the conditions and requirements of a project into design criteria, which are usually composed of topological relations among design elements, in order to explore the geometric properties of spaces and building components by means of generated 3D models.