Translation Quality
Evaluating machine translation (MT) quality focuses on assessing the accuracy, fluency, and overall naturalness of translated text, often comparing machine-generated translations to human references or using automatic metrics. Current research emphasizes improving MT quality through techniques like retrieval-augmented generation, preference-based alignment using LLMs (e.g., reinforcement learning from human feedback), and multi-agent collaborative approaches for complex texts. These advancements are crucial for enhancing cross-lingual communication and enabling broader access to information, impacting fields ranging from e-commerce to scientific research.
Papers
METEOR: Evolutionary Journey of Large Language Models from Guidance to Self-Growth
Jiawei Li, Chong Feng, Yang Gao
Transcending Language Boundaries: Harnessing LLMs for Low-Resource Language Translation
Peng Shu, Junhao Chen, Zhengliang Liu, Hui Wang, Zihao Wu, Tianyang Zhong, Yiwei Li, Huaqin Zhao, Hanqi Jiang, Yi Pan, Yifan Zhou, Constance Owl, Xiaoming Zhai, Ninghao Liu, Claudio Saunt, Tianming Liu
Beyond Human-Only: Evaluating Human-Machine Collaboration for Collecting High-Quality Translation Data
Zhongtao Liu, Parker Riley, Daniel Deutsch, Alison Lui, Mengmeng Niu, Apu Shah, Markus Freitag
Watching the Watchers: Exposing Gender Disparities in Machine Translation Quality Estimation
Emmanouil Zaranis, Giuseppe Attanasio, Sweta Agrawal, André F.T. Martins