Decoding Machine Translationese in English-Chinese News: LLMs vs. NMTs
Published in MT Summit 2025 (Technical Track), arXiv preprint, 2025
This study explores Machine Translationese (MTese) in English-to-Chinese news translation using classification and clustering with a five-layer feature set. It reveals distinguishable patterns between LLM and NMT outputs—such as sentence length, conjunction use, and lexical diversity—and confirms the detectability of MTese against original Chinese texts.
Recommended citation: Kong, D., & Macken, L. (2025). “Decoding Machine Translationese in English-Chinese News: LLMs vs. NMTs.” MT Summit 2025 (Technical Track), arXiv preprint arXiv:2506.22050.
Recommended citation: Kong, D., & Macken, L. (2025). "Decoding Machine Translationese in English-Chinese News: LLMs vs. NMTs." MT Summit 2025 (Technical Track), arXiv preprint arXiv:2506.22050.
Download Paper