Transition-Based Dependency Parsing using Perceptron Learner

TL;DR

This paper introduces a transition-based dependency parser trained with a Perceptron Learner, enhancing feature sets to outperform baseline models and state-of-the-art parsers in unlabeled attachment score.

Contribution

It presents a novel transition-based dependency parsing approach using an enhanced Perceptron Learner that surpasses existing models in accuracy.

Findings

Outperforms baseline arc-standard parser

Achieves higher UAS than MALT and LSTM parsers

Suggests methods for non-projective tree parsing

Abstract

Syntactic parsing using dependency structures has become a standard technique in natural language processing with many different parsing models, in particular data-driven models that can be trained on syntactically annotated corpora. In this paper, we tackle transition-based dependency parsing using a Perceptron Learner. Our proposed model, which adds more relevant features to the Perceptron Learner, outperforms a baseline arc-standard parser. We beat the UAS of the MALT and LSTM parsers. We also give possible ways to address parsing of non-projective trees.