Natural Language Understanding Wiki

Parse correction: Hall and Novák (2005)[1], Anguiano and Candito (2011)[2], Cetinoglu et al. (2011)[3]

Anguiano (2013)[4]: "We find that multiple-candidate approaches slightly improve parsing accuracy overall as well as for prepositional phrase attachment and coordination, two linguistic phenomena that exhibit high syntactic ambiguity."

Entity linking: Heinzerling et al. (2017)[5]


  1. Hall, Keith, and Václav Novák. "Corrective modeling for non-projective dependency parsing." In Proceedings of the Ninth International Workshop on Parsing Technology, pp. 42-52. Association for Computational Linguistics, 2005.
  2. Anguiano, E.H. and Candito, M., 2011, July. Parse correction with specialized models for difficult attachment types. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (pp. 1222-1233). Association for Computational Linguistics.
  3. Cetinoglu, Ozlem, Anton Bryl, Jennifer Foster, and Josef Van Genabith. "Improving dependency label accuracy using statistical post-editing: A cross-framework study." (2011).
  4. Anguiano, Enrique Henestroza. "Efficient large-context dependency parsing and correction with distributional lexical resources." PhD diss., Université Paris-Diderot-Paris VII, 2013.
  5. Heinzerling, B., Strube, M., & Lin, C.-Y. (2017). Trust, but Verify! Better Entity Linking through Automatic Verification. In EACL 2017 (pp. 828–838).