/mtg

Statistical discontinuous constituent parsing

Primary LanguageC++

mind_the_gap_v1.2: code for Maximin Coavoux, Benoît Crabbé. Représentation et analyse automatique des discontinuités syntaxiques dans les corpus arborés en constituants du français. TALN 2017 and for the experiments presented in my dissertation (forthcoming).

  • Lexicalized and unlexicalized structure-label transition systems
  • Joint constituency parsing, functional labelling and morphological analysis.
  • State-of-the-art pretrained models for
    • discontinuous constituency parsing: French, English (DPTB), German (Negra, Tiger)
    • projective constituency parsing: Arabic, Basque, French, German, Hebrew, Hungarian, Korean, Polish, Swedish (SPMRL)

Older versions: