CLaC at SemEval-2020 Task 5: Muli-task Stacked Bi-LSTMs

MinGyou Sung, Parsa Bagherzadeh, Sabine Bergler


Abstract
We consider detection of the span of antecedents and consequents in argumentative prose a structural, grammatical task. Our system comprises a set of stacked Bi-LSTMs trained on two complementary linguistic annotations. We explore the effectiveness of grammatical features (POS and clause type) through ablation. The reported experiments suggest that a multi-task learning approach using this external, grammatical knowledge is useful for detecting the extent of antecedents and consequents and performs nearly as well without the use of word embeddings.
Anthology ID:
2020.semeval-1.54
Volume:
Proceedings of the Fourteenth Workshop on Semantic Evaluation
Month:
December
Year:
2020
Address:
Barcelona (online)
Editors:
Aurelie Herbelot, Xiaodan Zhu, Alexis Palmer, Nathan Schneider, Jonathan May, Ekaterina Shutova
Venue:
SemEval
SIG:
SIGLEX
Publisher:
International Committee for Computational Linguistics
Note:
Pages:
445–450
Language:
URL:
https://aclanthology.org/2020.semeval-1.54
DOI:
10.18653/v1/2020.semeval-1.54
Bibkey:
Cite (ACL):
MinGyou Sung, Parsa Bagherzadeh, and Sabine Bergler. 2020. CLaC at SemEval-2020 Task 5: Muli-task Stacked Bi-LSTMs. In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 445–450, Barcelona (online). International Committee for Computational Linguistics.
Cite (Informal):
CLaC at SemEval-2020 Task 5: Muli-task Stacked Bi-LSTMs (Sung et al., SemEval 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.semeval-1.54.pdf