Italian immigration law is characterized by high normative density, layered administrative procedures, frequent cross-references, and significant practical consequences for non-EU citizens and public authorities. These features make it a challenging domain for legal formalization and for the development of reusable legal knowledge structures. This paper presents a first step towards a human-centered and LLM-assisted workflow for legal ontology design in public administration, using Article 5 of the Italian Immigration Act (Testo Unico dell’Immigrazione, TUI), on the residence permit, as an exploratory use case. Legal ontologies are framed not only as knowledge-representation artifacts, but also as legal-design artifacts: they restructure complex legal provisions into explicit concepts, relations, constraints, and procedural paths that can be inspected by humans and processed by machines. However, ontology construction remains labour-intensive and requires legal interpretation, domain expertise, and knowledge-engineering skills. Recent work has shown that LLMs can generate promising ontology drafts, while also producing omissions, superfluous elements, incorrect domains and ranges, and variable modeling quality. The proposed workflow treats LLM-generated artifacts as accountable drafts rather than authoritative outputs. Candidate OWL and SHACL artifacts are generated through structured prompting and then examined through a first Human-in-the-Loop (HITL) artifact inspection prototype. The prototype supports visualization, manual manipulation, and approval/rejection of generated fragments, as well as a graph-based legal design view. The contribution is not end-to-end automation of legal ontology engineering, but a first step towards making LLM-generated legal artifacts inspectable, editable, and accountable before reuse in public administration contexts.

Towards LLM-Assisted Legal Ontology Design for Public Administration: The Italian TUI Use Case

Said Daoudagh
;
2026

Abstract

Italian immigration law is characterized by high normative density, layered administrative procedures, frequent cross-references, and significant practical consequences for non-EU citizens and public authorities. These features make it a challenging domain for legal formalization and for the development of reusable legal knowledge structures. This paper presents a first step towards a human-centered and LLM-assisted workflow for legal ontology design in public administration, using Article 5 of the Italian Immigration Act (Testo Unico dell’Immigrazione, TUI), on the residence permit, as an exploratory use case. Legal ontologies are framed not only as knowledge-representation artifacts, but also as legal-design artifacts: they restructure complex legal provisions into explicit concepts, relations, constraints, and procedural paths that can be inspected by humans and processed by machines. However, ontology construction remains labour-intensive and requires legal interpretation, domain expertise, and knowledge-engineering skills. Recent work has shown that LLMs can generate promising ontology drafts, while also producing omissions, superfluous elements, incorrect domains and ranges, and variable modeling quality. The proposed workflow treats LLM-generated artifacts as accountable drafts rather than authoritative outputs. Candidate OWL and SHACL artifacts are generated through structured prompting and then examined through a first Human-in-the-Loop (HITL) artifact inspection prototype. The prototype supports visualization, manual manipulation, and approval/rejection of generated fragments, as well as a graph-based legal design view. The contribution is not end-to-end automation of legal ontology engineering, but a first step towards making LLM-generated legal artifacts inspectable, editable, and accountable before reuse in public administration contexts.
2026
Istituto di Scienza e Tecnologie dell'Informazione "Alessandro Faedo" - ISTI
LLM-Assisted Ontology Engineering, Human-in-the-Loop, Legal Design, Italian Immigration Act, Public Administration
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14243/608024
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ente

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact