An improved parser for data-oriented lexical-functional analysis

TL;DR

This paper introduces an improved LFG-DOP parser that leverages sentence fragments for parsing, demonstrating faster search methods, confirming the impact of fragment size on accuracy, and outperforming previous models in structure-based evaluations.

Contribution

The paper presents an enhanced LFG-DOP parser with faster search algorithms and confirms the benefits of larger fragments, outperforming previous models in structure-based parsing.

Findings

Viterbi n best search is 100 times faster than Monte Carlo search with same accuracy.

Parse accuracy increases with larger fragment size, confirming the DOP hypothesis.

LFG-DOP outperforms Tree-DOP in structure-only evaluations.

Abstract

We present an LFG-DOP parser which uses fragments from LFG-annotated sentences to parse new sentences. Experiments with the Verbmobil and Homecentre corpora show that (1) Viterbi n best search performs about 100 times faster than Monte Carlo search while both achieve the same accuracy; (2) the DOP hypothesis which states that parse accuracy increases with increasing fragment size is confirmed for LFG-DOP; (3) LFG-DOP's relative frequency estimator performs worse than a discounted frequency estimator; and (4) LFG-DOP significantly outperforms Tree-DOP is evaluated on tree structures only.