Extracting Normative Content from Legal Texts

Interoperability is a crucial issue, as large scale applications mainly depend on the possibility of mapping and connecting different models and structures. In the context of public services organization, achieving interoperability among public sector information means, on the one hand (back office), enabling dialogue among public offices, sharing data, standardising structures and models, connecting data producers and final users. On the other hand (front office), it means, providing better services and making citizens communicate and understand administrative and legal content. Due to the social dimension of legal information, semantic interoperability in the regulative domain is of crucial relevance, but of great complexity. Connecting regulative data requires the alignment of the structural layers, by adopting Semantic Web Standards, enabling data and meta-data to be identified, combined and retrieved, but also interconnection at conceptual level, allowing deeper communication of legal knowledge across domains and languages. The subject of this article concerns the conceptual modelling of legal information in systems for the organisation of digital services by administration and businesses and for assessing their normative compliance. This is a field in growing evolution and of great interest both from the market and public perspective, where several projects, tools and languages have been implemented, but the "knowledge bottleneck" in extracting and representing normative content from legal texts has not yet found a suitable solution, able to make the process of formalisation computationally tractable and reliable. The methodological aim is to show how the ontology-based approach, constructed around the basic notion of 'obligation', could support the formalisation of rules from texts and, at the same time, provide the reference model where to ground the mapping process between rules extracted from a normative statement and the business model specifications.