Through the COLaF project (Corpus et Outils pour les Langues de France, Corpus and Tools for the Languages of France), Inria aims to contribute to the development of free corpora and tools for French and other languages of France, in close collaboration with academic and institutional partners.

The scope of COLaF includes both:

• Text data (ALMAnaCH Inria Paris Centre),
• Speech and sign language data (MULTISPEECH, Inria Centre at the University of Lorraine).

COLaF aims to cover French and the languages of France in all its diversity:

• It aims to have a coverage as diverse as possible: French from France and elsewhere, regional languages, French-based creoles (including outside France), indigenous languages, migrant languages, French sign language.
• All aspects of variation will be studied, beyond the standard state of the language, including specialised languages, diachrony, non-standard states (user-generated content, learner language, etc.).

Activity within the project notably covers the acquisition and structuring of texts from non-textual sources (books, audio recordings, etc.), the classification by language and linguistic variety of large volumes of texts (in close connection with the OSCAR project), the development of annotation and transformation models (translation, normalisation, voice synthesis, sign language generation) serving the development of corpora and the exploitation of newly created resources.

COLaF is an Inria DEFI led by Benoît Sagot (head of the ALMAnaCH project team) and Slim Ouni (head of the MULTISPEECH project team).