conftrace_

2020 ACL ACL 2020

Efficient EUD Parsing

Abstract

AbstractWe present the system submission from the FASTPARSE team for the EUD Shared Task at IWPT 2020. We engaged with the task by focusing on efficiency. For this we considered training costs and inference efficiency. Our models are a combination of distilled neural dependency parsers and a rule-based system that projects UD trees into EUD graphs. We obtained an average ELAS of 74.04 for our official submission, ranking 4th overall.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Computer Science and Natural Language Processing

🐣 Hot Topic Early Bird - inference efficiency

Authors

Mathieu Dehouck, Mark Anderson, Carlos Gómez-Rodríguez

Topics

Machine Learning > Application Areas > Efficient Computing Computer Science > Applications > Document Analysis Deep Learning > Optimization & Theory > Model Compression Natural Language Processing > Applications > Semantic Parsing

Keywords

model distillation dependency parsing inference efficiency rule-based system neural dependency parser enhanced universal dependencies

More from ACL 2020

All ACL 2020 papers →

Turku Enhanced Parser Pipeline: From Raw Text to Enhanced Graphs in the IWPT 2020 Shared Task

Hybrid Enhanced Universal Dependencies Parsing

Adaptation of Multilingual Transformer Encoder for Robust Enhanced Universal Dependency Parsing

Linear Neural Parsing and Hybrid Enhancement for Enhanced Universal Dependencies

Enhanced Universal Dependency Parsing with Second-Order Inference and Mixture of Training Data

How Much of Enhanced UD Is Contained in UD?