Review for NeurIPS paper: Data Diversification: A Simple Strategy For Neural Machine Translation

Weaknesses: While the described approach is simple and very generally applicable, there are some major issues with the evaluation that need to be addressed. If 1. and 2. are addressed I would be willing to update my scores. The BLEU evaluation is not clearly described for the WMT and IWSLT experiments. Given the major variations observed in BLEU scores due to differences in post-processing or the BLEU evaluation script used, it's hard to fairly compare against previous work without clearly describing the post-processing, tokenization and BLEU evaluation tool used for these experiments. Since the proposed method relies heavily on using backward and forward translated data, these effects are bound to affect the observed BLEU improvements.