papers AI Learner
The Github is limit! Click to go to the new site.

Synthetic and Natural Noise Both Break Neural Machine Translation

2018-02-24
Yonatan Belinkov, Yonatan Bisk

Abstract

Character-based neural machine translation (NMT) models alleviate out-of-vocabulary issues, learn morphology, and move us closer to completely end-to-end translation systems. Unfortunately, they are also very brittle and easily falter when presented with noisy data. In this paper, we confront NMT models with synthetic and natural sources of noise. We find that state-of-the-art models fail to translate even moderately noisy texts that humans have no trouble comprehending. We explore two approaches to increase model robustness: structure-invariant word representations and robust training on noisy texts. We find that a model based on a character convolutional neural network is able to simultaneously learn representations robust to multiple kinds of noise.

Abstract (translated by Google)
URL

https://arxiv.org/abs/1711.02173

PDF

https://arxiv.org/pdf/1711.02173


Similar Posts

Comments