AI to better access and retrieve Judicial Interpretative Formulas – POLINE Final Conference

The event will feature the presentation of the POLINE tool and explore future applications of automated and open-access techniques for legal knowledge extraction, capable of improving the accessibility of judicial knowledge for the legal community.

The POLINE project has developed an AI-powered pilot tool for the retrieval and analysis of judicial interpretative formulas (JIFs) in the case-law of the Court of Justice of the European Union (CJEU) and selected national courts in the field of Value Added Tax (VAT).

The tool employs advanced AI, machine learning, and natural language processing (NLP) techniques to extract, cluster, and link JIFs. It is integrated into a modular platform composed of a Legal Database, Link Visualization, and a Customised Detection Module. Covering case-law from the CJEU as well as the Italian, Swedish, and Bulgarian Supreme Courts, the platform is designed for use by judges, legal practitioners, tax policymakers, and taxpayers.

Grounded in a multidisciplinary approach, POLINE combines theoretical and practical perspectives on judicial decision-making with legal informatics methods and AI-based analytical tools.

The project integrates the study of judicial interpretative formulas, the development of ontologies for VAT-related judicial concepts, and the creation of manually annotated training datasets to enhance the accuracy of machine learning models. Following a series of national trials and dissemination events, this final international conference will reflect on the experience and insights gained through the POLINE project. Discussions will highlight the importance of JIFs and their role in machine learning and NLP, drawing parallels with the function of JIFs in AI-assisted reasoning.