Abstract
Dependency parsers are almost ubiquitously evaluated on their accuracy scores, these scores say nothing of the complexity and usefulness of the resulting structures. As dependency parses are basic structures in which other systems are built upon, it would seem more reasonable to judge these parsers down the NLP pipeline.
In this chapter, we will discuss how different forms and different hybrid combinations of dependency parses effect the overall output of Syntax-Based machine translation both through automatic and manual evaluation. We show results from a variety of individual parsers, including dependency and constituent parsers, and describe multiple ensemble parsing techniques with their overall effect on the Machine Translation system.
We show that parsers’ UAS scores are more correlated to the NIST evaluation metric than to the BLEU Metric, however we see increases in both metrics. To truly see the effect of hybrid dependency parsers on machine translation, we will describe and evaluate a combine