This paper rethinks translation memory augmented neural machine translation (TM-augmented NMT) from two perspectives, i.e., a probabilistic view of retrieval and the variance-bias decomposition principle. The finding demonstrates that TM-augmented NMT is good at the ability of fitting data (i.e., lower bias) but is more sensitive to the fluctuations in the training data (i.e., higher variance), which provides an explanation to a recently reported contradictory phenomenon on the same translation task: TM-augmented NMT substantially advances vanilla NMT under the high-resource scenario whereas it fails under the low-resource scenario. Then we propose a simple yet effective TM-augmented NMT model to promote the variance and address the contradictory phenomenon. Extensive experiments show that the proposed TM-augmented NMT achieves consistent gains over both conventional NMT and existing TM-augmented NMT under two variance-preferable (low-resource and plug-and-play) scenarios as well as the high-resource scenario.

翻译：本文从两个角度重新审视了翻译记忆增强的神经机器翻译（TM-augmented NMT），即检索的概率视角和方差-偏差分解原理。研究发现表明，TM-augmented NMT在数据拟合能力上表现优异（即偏差较低），但对训练数据中的波动更为敏感（即方差较高），这为同一翻译任务中近期报告的一个矛盾现象提供了解释：在高资源场景下，TM-augmented NMT显著优于普通NMT，而在低资源场景下却表现不佳。随后，我们提出了一种简单但有效的TM-augmented NMT模型，以降低方差并解决这一矛盾现象。大量实验表明，所提出的TM-augmented NMT在两种方差占优场景（低资源和即插即用）以及高资源场景下，均比传统NMT和现有TM-augmented NMT取得了一致的性能提升。