Search papers, labs, and topics across Lattice.
Universit茅 C么te d'Azur, Inria
2
0
4
The integration of LLMs in ontology engineering reveals that fewer than 20% of extracted triples introduce unseen properties, challenging assumptions about legal data completeness.
Relying solely on Small Language Models, this approach achieved a surprising 27.80% content score while reducing reasoning bias.
