Analyzing Neural MT Search and Model Performance
This work addresses efficiency and performance in NMT systems, but it is incremental as it analyzes existing methods rather than introducing new ones.
The paper investigates whether more complex search algorithms or models are needed for neural machine translation (NMT), finding that current search algorithms are sufficient and that a small n-best list of 50 hypotheses contains notably better translations.
In this paper, we offer an in-depth analysis about the modeling and search performance. We address the question if a more complex search algorithm is necessary. Furthermore, we investigate the question if more complex models which might only be applicable during rescoring are promising. By separating the search space and the modeling using $n$-best list reranking, we analyze the influence of both parts of an NMT system independently. By comparing differently performing NMT systems, we show that the better translation is already in the search space of the translation systems with less performance. This results indicate that the current search algorithms are sufficient for the NMT systems. Furthermore, we could show that even a relatively small $n$-best list of $50$ hypotheses already contain notably better translations.