Implicit Distortion and Fertility Models for Attention-based Encoder-Decoder NMT Model
This addresses alignment issues in neural machine translation, which is an incremental improvement for translation systems.
The paper tackled the problem of incorrect alignment damaging translation quality in attention-based encoder-decoder neural machine translation models by attributing it to a lack of distortion and fertility models, and proposed new variations that achieved an improvement of 2 BLEU points over the original model.
Neural machine translation has shown very promising results lately. Most NMT models follow the encoder-decoder framework. To make encoder-decoder models more flexible, attention mechanism was introduced to machine translation and also other tasks like speech recognition and image captioning. We observe that the quality of translation by attention-based encoder-decoder can be significantly damaged when the alignment is incorrect. We attribute these problems to the lack of distortion and fertility models. Aiming to resolve these problems, we propose new variations of attention-based encoder-decoder and compare them with other models on machine translation. Our proposed method achieved an improvement of 2 BLEU points over the original attention-based encoder-decoder.