Sites Inria

Version française

ALPAGE Research team

Large-scale deep linguistic processing

  • Leader : Benoît Sagot
  • Research center(s) : CRI de Paris
  • Field : Perception, Cognition and Interaction
  • Theme : Language, Speech and Audio
  • Partner(s) : Université Denis Diderot (Paris 7)
  • Collaborator(s) : U. PARIS 7 (UPD), INRIA

Team presentation

Development of large-scale deep parsing systems for French: parsing techniques, deep linguistic descriptions (formalisms, lexicons, grammars) at the syntactic level, extension to the semantic and discourse levels.

Research themes

Natural Language Processing (NLP) is a research field which involves at the same time computer science, linguistics, as well as stochastics and cognition. Its aim is the automatic understanding and generation of human languages, be it in textual or oral form. ALPAGE works mostly on automatic understanding of French texts, although it works also on automatic generation as well as on languages other than French, including English. The strength of the ALPAGE team is partly due to the complementarity of its members: numerous breakthroughs in NLP have only been possible thanks to a close collaboration between computer scientists and linguists, all being NLP specialists. ALPAGE, which is a joint project-team between INRIA Paris - Rocquencourt and the Department of Linguistics of Université Paris 7, aims to improve significantly the state-of-the-art in French parsing. This requires a better understanding and a better formalization of linguistic phenomena, including the most complex ones, followed by their integration in lexical and grammatical models, which, in turn, are implemented thanks to advanced algorithmic and engineering techniques, both symbolic and probabilistic, targeted towards lexical and syntactic analysis, but also automatic generation, and, in the long-term, automatic translation. Applications of NLP are numerous and important for ALPAGE, and namely spelling, grammar and style correction tools, information retrieval tools, and text mining tools.

Keywords: NLP Parsing Lexicons Discourse