DIRECT: Ontology-based Discovery of Responsibility and Causality in Legal Case Descriptions

In this paper we present DIRECT, a system for automatic discovery of responsibility and causal relations in legal case descriptions based on LRI-Core, a core ontology that covers the main concepts that are common to all legal domains. These domains have a predominant common-sense character – the law is still for the people – and typical legal concepts such as norm, role responsibility, contract, etc. have a grounding in abstract common-sense conceptualizations. A common sense framework is even more necessary for the automatic discovery of the causal structure of legal case descriptions, since they often are even devoid of a specialized legal conceptualization. The paper presents a number of design principles that follow from the common-sense stance in developing the LRI-Core: the most important being cognitive plausibility. Furthermore, an approach is presented to enable the automatic analysis of cases described in terms of the ontology. Such analysis will determine causal chains in situation descriptions on the basis of which responsibility attribution can take place.