Law Enforcement Ontology for Identification of Related Information of Interest Across Free Text Dcouments

A law enforcement ontology that incorporates extensions such as Thesauri, specialized rules, abductive hypothesis and process modeling for expansion of extracted entity phrases, is described. The ontology is part of a project to facilitate automated, reliable identification of related information of interest found in law enforcement-related free-text documents. Results of testing on a complex, real-world law enforcement dataset show that the addition of the ontology significantly improves the expanded entity phrase extraction used for the identification of related information of interest in free-text documents and merits additional expansion. Future work will add semantic inference and insertion functions and extend the specialized rules and abductive hypotheses components.