Séminaire Cognition & Langage

The presentation will start with a definition, examples and an overview of the importance of “dead” (classical) languages and will move on to present the challenges that they are associated with with regards to NLP (scarcely resourced; limited genres and registers). The discussion will then move onto a presentation of concrete examples of NLP techniques as applied to “dead” languages. Firstly, common tools related to these languages will be introduced (e.g. tokenizers, POS taggers). Then, examples of research projects will be presented – starting from projects with large global impact (e.g. finetuning of LLMs for classical languages, temporal studies of language development) and moving on to more narrowly-focused research that the presenter has participated in (automatic classification of the variants in witnesses of Biblical Hebrew texts; semi-automatic generation of a graded reader in Old Church Slavonic).