Hierarchical Pointer Net Parsing

TL;DR

This paper introduces hierarchical pointer network parsers that improve tree-structured parsing tasks by providing a better inductive bias, achieving state-of-the-art results on dependency and discourse parsing benchmarks.

Contribution

The paper proposes hierarchical pointer network parsers that enhance tree structure derivation, outperforming existing sequential decoder models in parsing tasks.

Findings

Outperforms existing methods on standard benchmarks

Sets new state-of-the-art in dependency parsing

Effective for sentence-level discourse parsing

Abstract

Transition-based top-down parsing with pointer networks has achieved state-of-the-art results in multiple parsing tasks, while having a linear time complexity. However, the decoder of these parsers has a sequential structure, which does not yield the most appropriate inductive bias for deriving tree structures. In this paper, we propose hierarchical pointer network parsers, and apply them to dependency and sentence-level discourse parsing tasks. Our results on standard benchmark datasets demonstrate the effectiveness of our approach, outperforming existing methods and setting a new state-of-the-art.