On the Complementarity between Pre-Training and Back-Translation for Neural Machine Translation
It addresses the problem of effectively utilizing monolingual data for machine translation, but is incremental as it builds on existing PT and BT methods.
This paper investigates the complementarity between pre-training (PT) and back-translation (BT) for neural machine translation, finding that PT mainly benefits the encoder and BT the decoder, leading to state-of-the-art performances on WMT16 English-Romanian and English-Russian benchmarks.
Pre-training (PT) and back-translation (BT) are two simple and powerful methods to utilize monolingual data for improving the model performance of neural machine translation (NMT). This paper takes the first step to investigate the complementarity between PT and BT. We introduce two probing tasks for PT and BT respectively and find that PT mainly contributes to the encoder module while BT brings more benefits to the decoder. Experimental results show that PT and BT are nicely complementary to each other, establishing state-of-the-art performances on the WMT16 English-Romanian and English-Russian benchmarks. Through extensive analyses on sentence originality and word frequency, we also demonstrate that combining Tagged BT with PT is more helpful to their complementarity, leading to better translation quality. Source code is freely available at https://github.com/SunbowLiu/PTvsBT.