Universal Dependency Parsing from Scratch
This addresses the challenge of building an end-to-end dependency parsing system for NLP researchers and practitioners, though it is incremental as it builds on existing shared task frameworks.
The paper tackles the problem of universal dependency parsing from raw text, presenting a complete neural pipeline system that achieved competitive performance on big treebanks and would have ranked 2nd, 1st, and 3rd on official metrics after a bug fix.
This paper describes Stanford's system at the CoNLL 2018 UD Shared Task. We introduce a complete neural pipeline system that takes raw text as input, and performs all tasks required by the shared task, ranging from tokenization and sentence segmentation, to POS tagging and dependency parsing. Our single system submission achieved very competitive performance on big treebanks. Moreover, after fixing an unfortunate bug, our corrected system would have placed the 2nd, 1st, and 3rd on the official evaluation metrics LAS,MLAS, and BLEX, and would have outperformed all submission systems on low-resource treebank categories on all metrics by a large margin. We further show the effectiveness of different model components through extensive ablation studies.