Developing a topological information extraction model for space syntax analysis

Computer-aided design (CAD) systems have evolved into more intelligent systems, employing building information modeling (BIM) to manage semantic building information including the entities of building components and spaces. Based on the technological advancement, there have been some researches for conducting spatial analysis in CAD or simulation systems. However, in the previous works, the task to subdivide a concave space into convex spaces has been carried out manually, which caused inefficiency of time and cost. Therefore, this study has intended to develop a topological information extraction model (TIEM) to extract topological information, recognize geometric and topological features of spaces, and select convex spaces best fit for Space Syntax analysis in an automatic manner. To reach this goal, this study developed algorithms to extract essential information from semantic building information including building components and spaces produced in the form of Industry Foundation Classes (IFC), and employed graph theory and binary spatial partitioning (BSP) tree to recognize spaces and decompose a concave space into optimal convex spaces. As a result, this study has successfully decomposed a concave space into convex spaces automatically, has drawn a j-graph with the divided convex unit spaces, and has measured spatial configurations based on Space Syntax theory indentifying social and spatial properties of a built environment. These results indicate that the TIEM will serve as useful means to help architects to find an appropriate purpose of each space for sustainable built environment on the basis of the spatial and social relationships in buildings or urban systems.