Séminaire ayant lieu dans le cadre du doctorat en informatique cognitive, en collaboration avec le centre de recherche CRIA et l'ISC     

Roberto NAVIGLI

Jeudi le 10 avril 2025 à 10h30

Local: PK-5115   (Il est possible d'y assister en virtuel en vous inscrivant ici)   

TITRE :  Can Large Language Models Prove Their Understanding of Language?

RÉSUMÉ

I will discuss some developments in multilingual lexical semantics and word sense disambiguation (WSD). BabelNet (Navigli & Ponzetto, 2012) introduced a large-scale, automatically constructed multilingual semantic network, integrating structured and unstructured lexical resources to support cross-lingual applications. A 2009 survey provided a comprehensive analysis of WSD methodologies, highlighting the challenges of ambiguity resolution and the evolution of knowledge-based and statistical approaches. A more recent survey (Bevilacqua et al., 2021) tracks developments in WSD, emphasizing neural architectures and data-driven improvements. These works have helped shape the understanding of semantic representation and disambiguation in Natural Language Processing.

BIOGRAPHIE

Roberto NAVIGLI, is Professor in the Department of Computer, Control and Management Engineering at Sapienza University of Rome, where he leads the Sapienza Natural Language Processing (NLP) Group. His research focuses on multilingual NLP, computational semantics, and knowledge representation. He developed BabelNet, a multilingual lexical-semantic knowledge graph that integrates resources like WordNet, Wikipedia, and Wiktionary and has contributed to word sense disambiguation, creating large-scale, automatically extracted training sets. In semantic role labeling, he has highlighted the need for improved models to handle diverse non-verb predicate types, such as nouns and adjectives, and he has contributed to multilingual semantic parsing techniques for creating language-independent semantic representations. 

RÉFÉRENCES

Bevilacqua, M., Pasini, T., Raganato, A., & Navigli, R. (2021). Recent trends in word sense disambiguation: A survey. International Joint Conference on Artificial Intelligence (pp. 4330-4338).

Navigli, R., & Ponzetto, S. P. (2012). BabelNet: The automatic construction, evaluation and application of a wide-coverage multilingual semantic network. Artificial Intelligence, 193, 217-250.

Navigli, R. (2009). Word sense disambiguation: A survey. ACM computing surveys (CSUR), 41(2), 1-69.