Most translation tasks among languages belong to the zero-resource translation problem where parallel corpora are unavailable. Multilingual neural machine translation (MNMT) enables one-pass translation using shared semantic space for all languages compared to the two-pass pivot translation but often underperforms the pivot-based method. In this paper, we propose a novel method, named as Unified Multilingual Multiple teacher-student Model for NMT (UM4). Our method unifies source-teacher, target-teacher, and pivot-teacher models to guide the student model for the zero-resource translation. The source teacher and target teacher force the student to learn the direct source-target translation by the distilled knowledge on both source and target sides. The monolingual corpus is further leveraged by the pivot-teacher model to enhance the student model. Experimental results demonstrate that our model of 72 directions significantly outperforms previous methods on the WMT benchmark.
@inproceedings{um4,
title = {UM4: Unified Multilingual Multiple Teacher-Student Model for Zero-Resource Neural Machine Translation},
author = {Yang, Jian and Yin, Yuwei and Ma, Shuming and Zhang, Dongdong and Wu, Shuangzhi and Guo, Hongcheng and Li, Zhoujun and Wei, Furu},
booktitle = {Proceedings of the Thirty-First International Joint Conference on
Artificial Intelligence, {IJCAI-22}},
publisher = {International Joint Conferences on Artificial Intelligence Organization},
editor = {Lud De Raedt},
pages = {4454--4460},
year = {2022},
month = {7},
note = {Main Track},
doi = {10.24963/ijcai.2022/618},
url = {https://doi.org/10.24963/ijcai.2022/618},
}