# Multi-Task Semantic Dependency Parsing with Policy Gradient for Learning Easy-First Strategies

arXiv: https://arxiv.org/abs/1906.01239 · HTML: https://mlsanity.com/paper/1906.01239/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Jun 2019.

## Built on by (most cited first, 9 of 9)

- [Multi-Task Learning in Natural Language Processing: An Overview](https://mlsanity.com/paper/2109.09138/) (2021)
- [End-to-end Semantic Role Labeling with Neural Transition-based Model](https://mlsanity.com/paper/2101.00394/) (2021)
- [A Conditional Splitting Framework for Efficient Constituency Parsing](https://mlsanity.com/paper/2106.15760/) (2021)
- [Revisiting Conversation Discourse for Dialogue Disentanglement](https://mlsanity.com/paper/2306.03975/) (2023)
- [Modeling Label Correlations for Second-Order Semantic Dependency Parsing with Mean-Field Inference](https://mlsanity.com/paper/2204.03619/) (2022)
- [Dependency Parsing with Backtracking using Deep Reinforcement Learning](https://mlsanity.com/paper/2206.13914/) (2022)
- [Policy Optimization Prefers The Path of Least Resistance](https://mlsanity.com/paper/2510.21853/) (2025)
- [A Higher-Order Semantic Dependency Parser](https://mlsanity.com/paper/2201.11312/) (2022)
- [Auxiliary Tasks to Boost Biaffine Semantic Dependency Parsing](https://mlsanity.com/paper/2402.07682/) (2024)

## Abstract (arXiv)

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.