Lexicographie informatique et intelligence artificielle : analyse contrastive des évolutions en France et en Chine

Abstract

This article offers a comparative study of the development of computational lexicography in France and China, analyzing their historical trajectories, institutional frameworks, and technological orientations. In France, the discipline is rooted in an academic tradition structured around major infrastructures such as the ATILF, the TLFi, Frantext, and the CNRTL, with a strong emphasis on digital humanities, data standardization (XML/TEI, LLOD), and interoperability, particularly through tools and models such as NooJ and OntoLex-Lemon. In China, development, although more recent, has been rapid since the 1980s and is based on the creation of massive corpora, semantic databases such as HowNet, and the increasing integration of artificial intelligence and large language models into lexicographical production. The comparison highlights two complementary approaches: a heritage-based and methodologically rigorous tradition in France, and a pragmatic, technology-driven dynamic in China, paving the way for multilingual, and interconnected digital lexicography.