Search

This paper outlines the ParlaMint project from the perspective of its goals, tasks, participants, results and applications potential. The project produced language corpora from the sessions of the national parliaments of 17 countries, almost half a billion words in total. The corpora are split into COVID-related subcorpora (from November 2019) and reference corpora (to October 2019). The corpora are uniformly encoded according to the ParlaMint schema with the same Universal Dependencies linguistic annotations. Samples of the corpora and conversion scripts are available from the project’s GitHub repository. The complete corpora are openly available via the CLARIN.SI repository1 for download, and through the NoSketch Engine2 and KonText3 concordancers as well as through the Parlameter4 interface for exploration and analysis.

Legal science is quite distinctive compared to other disciplines, also social sciences disciplines (variety of publications, interlink between practice and research, multiple audiences, etc.). In addition, the fact that legal science has always shied away from bibliometrics-based evaluation practices make the field a specific case study. The case study presented in this chapter is located precisely in this context and within a heated discussion that sees the law domain and its communities at the centre of the debate. An overview of the practices, policies, and criteria applied in law journal assessment in three countries (Croatia, Italy, and Spain) is provided. The panorama is varied, but it is constantly changing. Most of the qualitative criteria for journal assessment are common to all three countries, but significant differences are in place when it comes to mandating these criteria in practice. According to the survey on law journals across countries, an adequate definition of the appropriate criteria for assessing journal quality could significantly improve journals’ editorial policies and therefore the quality of legal publications.

The European Case Law Identifier (ECLI) was established in 2010. It has been implemented by three European courts and (partly or in full) by twelve Member States, while eight Member States are preparing an implementation. During these implementations various issues concerning the format, metadata and the architecture of ECLI have surfaced, while in the meantime technological insights have progressed. To improve the accessibility of current and future case law repositories an ECLI 2.0 needs to be considered. Given the implementations realized, backwards compatibility is an absolute prerequisite. In this paper the various issues and possible solutions are discussed.

Interoperability today is the key term for the enhancement of databases published on the web: the data, when isolated, have little value, on the contrary, their value increases significantly when different datasets, produced and published independently by different providers, can be reused and freely mashed by third parties. The use of data for new purposes not foreseen by organizations and individuals who publish "raw data" is one of the advantages of linked open data model. To achieve these benefits content and relationships between the entities described in the dataset should be explicitly represented in standard web formats (XML, RDF, URI). The DoGi-Legal Literature database, one of the most valuable sources for online access to legal doctrine, created and managed by the Institute of Legal Information Theory and Techniques of the CNR is following this direction. This paper will define the schema of the data representing the database in RDF format. This will make the DoGi database interoperable between different data and service providers (libraries, publishers, information services for accessing national and European legal information), allowing the creation of new advanced services to be made available on the web of data. In particular, the contribution will focus on the goal to promote semantic interoperability between the DoGi classification schema and other semantic indexing tools in legal domain.

The need for bridging the gap between linguistically-oriented knowledge resources (i.e. lexicons) and domain-oriented ones (i.e. ontologies) is acknowledged within both the NLP and the AI&Law community. In this paper we propose to face this need by comparing a FrameNet-style and an ontological characterization of the obligation Fundamental Legal Concept. In particular, we carried out a case-study aimed at investigating whether and to which extent different views on this Fundamental Legal Concept offered by the FrameNet resource can be mapped to an ontological characterization of the complex concept of public function, stemmed from the basic normative position obligation.