LEVERAGING GRAPH BASED SEMANTIC ENRICHMENT FOR ENHANCED AUTOMATED CODE-CHECKING

Despite advances in BIM technology, achieving a comprehensive machine-readable design representation for Automated Code Checking (ACC) remains a challenge. We propose decomposing regulatory checks into sets of semantic enrichment tasks, each tailored with a suitable solution. We demonstrate this approach with a case study on Israeli regulations for security rooms, focusing on two distinct semantic enrichment tasks. We use the Girvan-Newman algorithm to isolate single apartments in a floor plan. A GNN then categorizes security rooms based on the ratio of supporting walls, according to the ranges defined in the regulations. The approach paves the way for a highly automated ACC methodology, with potential implications for diverse regulatory contexts.