Simple and Accurate Dependency Parsing Using Bidirectional LSTM Feature Representations

TL;DR

This paper introduces a simple yet effective dependency parsing method using bidirectional LSTM features, achieving state-of-the-art accuracy on English and Chinese with minimal architecture complexity.

Contribution

The paper presents a novel approach that integrates BiLSTM-based feature representations into dependency parsers, improving accuracy with simpler models.

Findings

Achieves state-of-the-art accuracy on English and Chinese datasets.

Uses joint training of BiLSTM and parser for effective feature extraction.

Works with both greedy and globally optimized parsers.

Abstract

We present a simple and effective scheme for dependency parsing which is based on bidirectional-LSTMs (BiLSTMs). Each sentence token is associated with a BiLSTM vector representing the token in its sentential context, and feature vectors are constructed by concatenating a few BiLSTM vectors. The BiLSTM is trained jointly with the parser objective, resulting in very effective feature extractors for parsing. We demonstrate the effectiveness of the approach by applying it to a greedy transition-based parser as well as to a globally optimized graph-based parser. The resulting parsers have very simple architectures, and match or surpass the state-of-the-art accuracies on English and Chinese.