Turning Fixed to Adaptive: Integrating Post-Evaluation into Simultaneous Machine Translation
Shoutao Guo 1,2, Shaolei Zhang 1,2, Yang Feng 1,2 1Key Laboratory of Intelligent Information Processing Institute of Computing Technology, Chinese Academy of Sciences (ICT/CAS) 2 University of Chinese Academy of Sciences, Beijing, China {guoshoutao22z, zhangshaolei20z, Corresponding author: Yang Feng.
Abstract
Simultaneous machine translation (SiMT) starts its translation before reading the whole source sentence and employs either fixed or adaptive policy to generate the target sentence. Compared to the fixed policy, the adaptive policy achieves better latency-quality tradeoffs by adopting a flexible translation policy. If the policy can evaluate rationality before taking action, the probability of incorrect actions will also decrease. However, previous methods lack evaluation of actions before taking them. In this paper, we propose a method of performing the adaptive policy via integrating post-evaluation into the fixed policy. Specifically, whenever a candidate token is generated, our model will evaluate the rationality of the next action by measuring the change in the source content. Our model will then take different actions based on the evaluation results. Experiments on three translation tasks show that our method can exceed strong baselines under all latency111Code is available at https://github.com/ictnlp/PED-SiMT.
中文速览
同声翻译（Simultaneous Machine Translation）系统需要在没有读完完整源句子的情况下就开始输出译文，关键难题在于如何在恰当的时机决定"继续读入源词"还是"输出译词"，而以往方法都是先做决策再行动，缺乏对动作合理性的事前评估。本文提出一种将"事后评估"融入固定策略的自适应翻译策略（Post-Evaluation Policy, PED-SiMT）：每生成一个候选译词后，模型借助胶囊网络（Capsule Networks）计算各源词的"翻译程度"，再通过比较相邻步骤翻译程度的变化量来判断当前动作是否合理——变化显著则输出，否则继续读入源词。为使模型在训练阶段覆盖更多翻译路径，作者还设计了带扰动的多路径训练方式。在WMT15德→英、IWSLT15英→越及IWSLT14英→德三项翻译任务上，该方法在所有延迟档位下均超越了强基线，对于需要在低延迟与高质量之间取得平衡的实时翻译场景具有重要的实用价值。
原文 arXiv:2210.11900;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2210.11900v1