Improving Neural Parsing by Disentangling Model Combination and Reranking Effects
arXiv:1707.03058
Abstract
Recent work has proposed several generative neural models for constituency parsing that achieve state-of-the-art results. Since direct search in these generative models is difficult, they have primarily been used to rescore candidate outputs from base parsers in which decoding is more straightforward. We first present an algorithm for direct search in these generative models. We then demonstrate that the rescoring results are at least partly due to implicit model combination rather than reranking effects. Finally, we show that explicit model combination can improve performance even further, resulting in new state-of-the-art numbers on the PTB of 94.25 F1 when training only on gold data and 94.66 F1 when using external data.
arXiv commentACL 2017. The first two authors contributed equally
Cited by in corpus (5)
- Natural Language Processing Advancements By Deep Learning: A Survey
- Global Reasoning over Database Structures for Text-to-SQL Parsing
- Two Local Models for Neural Constituent Parsing
- Direct Output Connection for a High-Rank Language Model
- Policy Gradient as a Proxy for Dynamic Oracles in Constituency Parsing