Don't Stop Me Now! Using Global Dynamic Oracles to Correct Training Biases of Transition-Based Dependency Parsers

Lauriane Aufrant (1) , Guillaume Wisniewski (1) , François Yvon (1)

1 LIMSI - Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur

This paper formalizes a sound extension of dynamic oracles to global training, in the frame of transition-based dependency parsers. By dispensing with the pre-computation of references, this extension widens the training strategies that can be entertained for such parsers; we show this by revisiting two standard training procedures, early-update and max-violation, to correct some of their search space sampling biases. Experimentally, on the SPMRL treebanks, this improvement increases the similarity between the train and test distributions and yields performance improvements up to 0.7 UAS, without any computation overhead.

Mots clés: dependency parsing; transition-based system; machine learning

Domaines: Informatique et langage [cs.CL]

Lauriane Aufrant, Guillaume Wisniewski, François Yvon. Don't Stop Me Now! Using Global Dynamic Oracles to Correct Training Biases of Transition-Based Dependency Parsers. Conference of the European Chapter of the ACL, Jan 2017, Valencia, Spain. pp.318 - 323, ⟨10.18653/v1/E17-2051⟩. ⟨hal-01618377⟩