The Twenty-Third International Conference on Software Engineering and Knowledge Engineering
Our society’s reliance on a variety of critical infrastructures (CI) presents significant challenges for disaster preparedness, response and recovery. Experts from different domains including police, paramedics, firefighters and various other CI teams are involved in the fast paced response to a disaster, increasing the risk of miscommunication. To ensure clear communication, as well as to facilitate CI software interoperability, a common disaster ontology is needed. We propose using the knowledge stored in domain glossaries, vocabularies and dictionaries for the creation of a lightweight disaster management domain ontology. Glossaries, vocabularies and dictionaries are semi structured representations of domain knowledge, where significant human effort has been invested in choosing relevant terms, determining their definitions, acronyms, synonyms and sometimes even relations. We use that knowledge built into semi formatted documents for ontology learning. In particular, we look at five glossaries/vocabularies from the disaster management domain and analyze their content similarity and structure. A lightweight disaster ontology is created exploiting the structure of the semi-structured source documents.