Uses natural language processing (NLP) to extract structured concepts from chemical accident database narratives and automatically populate an ontology, enabling richer queries of causes, consequences, substances, and contextual factors. For PSM Management of Change, the practical value is “learning before changing”: MOC reviewers can retrieve analogous historical incidents for similar materials/equipment/operations and incorporate lessons into change risk reviews, safeguard requirements, and procedure/training updates. The approach also supports organizational memory, helping prevent repeat incidents during or after modifications.
