Towards Better UD Parsing: Deep Contextualized Word Embeddings, Ensemble, and Treebank Concatenation

Abstract

This paper describes our system (HIT-SCIR) submitted to the CoNLL 2018 shared task on Multilingual Parsing from Raw Text to Universal Dependencies. We base our submission on Stanford's winning system for the CoNLL 2017 shared task and make two effective extensions: 1) incorporating deep contextualized word embeddings into both the part of speech tagger and parser; 2) ensembling parsers trained with different initialization. We also explore different ways of concatenating treebanks for further improvements. Experimental results on the development data show the effectiveness of our methods. In the final evaluation, our system was ranked first according to LAS (75.84%) and outperformed the other systems by a large margin.

arXiv commentSystem description paper of our system (HIT-SCIR) for the CoNLL 2018 shared task on Universal Dependency parsing, which was ranked first in the LAS evaluation. Fix typos and grammar errors. Add the results of parser without ensemble

References in corpus (4)

• Enriching Word Vectors with Subword Information
• Neural Probabilistic Model for Non-projective MST Parsing
• Cross-lingual Word Segmentation and Morpheme Segmentation as Sequence Labelling
• Distilling Knowledge for Search-based Structured Prediction

Cited by in corpus (7)

• A review of sentiment analysis research in Arabic language
• Resources for Turkish Natural Language Processing: A critical survey
• Large-Scale Contextualised Language Modelling for Norwegian
• Constructive Type-Logical Supertagging with Self-Attention Networks
• Exploring Descriptions of Movement Through Geovisual Analytics
• The ADAPT Enhanced Dependency Parser at the IWPT 2020 Shared Task
• A Hybrid Approach to Dependency Parsing: Combining Rules and Morphology with Deep Learning