Application of Semantic Web to Incident Reporting

While lots of incident reports are accumulated, analyzing them is not kept up with at present. For this reason, a database of incident reports is not utilized sufficiently and reports are only accumulated. Under this situation, efficient way of handling such information is needed. For enabling computer processing of incident reports, a semantic Web approach is applied to incident reporting. The form for reporting is defined using XML and incident ontology is constructed. The expressions in incident reports are tagged referring to the ontology. Reasoning by intelligent agents based on the ontology is useful for analyzing background factors of problems or extracting lessons for the future. Some concrete examples are shown using actual incident reports from nuclear facilities. This approach makes a database more useful and worthwhile, and it allow using more energy for countermeasures rather than analysis. In this research, a conceptual design of applying semantic Web to incident reporting is proposed. The paper discusses the first stage of this application: definition of metadata tags and incident ontology, and tagging of existing data