Multitask Easy-First Dependency Parsing: Exploiting Complementarities of Different Dependency Representations

Yash Kankanampati (1) , Joseph Le Roux (2) , Nadi Tomeh (3) , Dima Taji (4) , Nizar Habash (4)

1 IIT DHANBAD - ISM - Indian Institute of Technology [Dhanbad]
2 UP13 - Université Paris 13
3 LIPN - Laboratoire d'Informatique de Paris-Nord
4 New York University [Abu Dhabi]

We present a parsing model for projective dependency trees which takes advantage of the existence of complementary dependency annotations for a language. This is the case for Arabic with the availability of CATiB and UD treebanks. Our system performs syntactic parsing according to both annotation types jointly as a sequence of arc-creating operations following the Easy-First approach, and partially created trees for one annotation type are also available to the other as features for the score function. This method gives error reduction of 9.9% on CATiB and 6.1% on UD compared to a single-task baseline, and ablation tests show that the main contribution of this reduction is given by sharing tree representation between tasks, and not simply sharing BiLSTM layers as is usually performed in NLP multitask systems.

Domains: Artificial Intelligence [cs.AI]; Document and Text Processing; Computer Science [cs]

Yash Kankanampati, Joseph Le Roux, Nadi Tomeh, Dima Taji, Nizar Habash. Multitask Easy-First Dependency Parsing: Exploiting Complementarities of Different Dependency Representations. 28th International Conference on Computational Linguistics, Dec 2020, Barcelona (on line), Spain. ⟨10.18653/v1/2020.coling-main.225⟩. ⟨hal-03168039⟩