Multi-Task Semantic Dependency Parsing with Policy Gradient for Learning Easy-First Strategies
Computation and Language
2019-06-05 v1
Abstract
In Semantic Dependency Parsing (SDP), semantic relations form directed acyclic graphs, rather than trees. We propose a new iterative predicate selection (IPS) algorithm for SDP. Our IPS algorithm combines the graph-based and transition-based parsing approaches in order to handle multiple semantic head words. We train the IPS model using a combination of multi-task learning and task-specific policy gradient training. Trained this way, IPS achieves a new state of the art on the SemEval 2015 Task 18 datasets. Furthermore, we observe that policy gradient training learns an easy-first strategy.
Cite
@article{arxiv.1906.01239,
title = {Multi-Task Semantic Dependency Parsing with Policy Gradient for Learning Easy-First Strategies},
author = {Shuhei Kurita and Anders Søgaard},
journal= {arXiv preprint arXiv:1906.01239},
year = {2019}
}
Comments
ACL2019 Long accepted. 9 pages for the paper and the additional 2 pages for the supplemental material