Synslator: An Interactive Machine Translation Tool with Online Learning
Jiayi Wang, Ke Wang, Fengming Zhou, Chengyu Wang, Zhiyong Fu, Zeyu Feng, Yu Qing Zhao, Yuqi Zhang · 2024
Interactive Machine Translation (IMT) advances the computer-aided translation (CAT) paradigm, enabling collaboration between machine translation systems and human translators for high-quality outputs. This paper presents Synslator, a CAT tool designed for IMT and proficient in online learning with real-time translation memories. Synslator accommodates different CAT service deployments by integrating two neural translation models for online learning and a language model to boost translation fluency interactively. Our evaluations demonstrate the system's online learning effectiveness, showing a 13% increase in post-editing efficiency with Synslator's interactive features. A tutorial video is provided at: https://youtu.be/K0vRsb2lTt8.