Effective Strategies in Zero-Shot Neural Machine Translation
This work addresses the problem of zero-resource translation for multilingual systems, but it appears incremental as it builds on existing methods.
The paper tackled zero-shot neural machine translation without parallel corpora by proposing two strategies, which improved performance and efficiency, particularly in multilingual translation with unbalanced data and alleviated language bias.
In this paper, we proposed two strategies which can be applied to a multilingual neural machine translation system in order to better tackle zero-shot scenarios despite not having any parallel corpus. The experiments show that they are effective in terms of both performance and computing resources, especially in multilingual translation of unbalanced data in real zero-resourced condition when they alleviate the language bias problem.