Splitting EUD Graphs into Trees: A Quick and Clatty Approach

Mark Anderson, Carlos Gómez-Rodríguez


Abstract
We present the system submission from the FASTPARSE team for the EUD Shared Task at IWPT 2021. We engaged in the task last year by focusing on efficiency. This year we have focused on experimenting with new ideas on a limited time budget. Our system is based on splitting the EUD graph into several trees, based on linguistic criteria. We predict these trees using a sequence-labelling parser and combine them into an EUD graph. The results were relatively poor, although not a total disaster and could probably be improved with some polishing of the system’s rough edges.
Anthology ID:
2021.iwpt-1.17
Volume:
Proceedings of the 17th International Conference on Parsing Technologies and the IWPT 2021 Shared Task on Parsing into Enhanced Universal Dependencies (IWPT 2021)
Month:
August
Year:
2021
Address:
Online
Venues:
ACL | IJCNLP | IWPT
SIG:
SIGPARSE
Publisher:
Association for Computational Linguistics
Note:
Pages:
167–174
Language:
URL:
https://aclanthology.org/2021.iwpt-1.17
DOI:
10.18653/v1/2021.iwpt-1.17
Bibkey:
Copy Citation:
PDF:
https://aclanthology.org/2021.iwpt-1.17.pdf