Ontology-Driven Automated Reasoning About Property Crimes

• DOI: 10.1007/s12599-024-00886-3
• Publication Date: 2024-08-12T14:03:36Z
• Authors: Francisco Navarrete; Ángel L. Garrido; Carlos Bobed; Manuel Atencia; Antonio Vallecillo
• Keywords: Automated reasoning

Abstract The classification of police reports according to the typification criminal act described in them is not an easy task. are written natural language and often present missing, imprecise, or even inconsistent information, lack sufficient details make a clear decision. Focusing on property crimes, aim this work assist judges process by automatically extracting information from producing list possible classifications crimes accompanied degree confidence each them. follows design science research methodology, developing tool as artifact. proposal uses extraction techniques obtain data reports, guided ontology developed for Spanish legal system crimes. Probabilistic inference mechanisms used select set articles law that could apply given case, when evidence does allow unambiguous identification. has been empirically validated real environment with prosecutors. results show feasible usable, be effective assisting classify crime reports.