Teaching Translation Verification with AI Corpora: A Case Study of Chinese Culture-Loaded Terms

Students of Chinese–English translation frequently render culture-loaded terms by pinyin transliteration or word-for-word literal translation, rarely verifying their choices against usage evidence. This paper reports a case study of integrating an AI corpus platform (AI-Corpora) into verification training for culture-loaded terms in an undergraduate translation course at an applied university in China. A five-step verification model was designed and implemented—independent drafting, corpus search, evidence evaluation, guided revision, and term archiving—across a teaching unit containing three illustrative tasks: a cured-meat term in a Sichuan food text, teahouse-culture terms, and solar-term names. Analysis of before-and-after student outputs shows that corpus evidence helps learners move from pinyin-only or literal renderings toward accepted and audience-appropriate translations, and fosters a data-driven habit of translation decision-making. The paper further discusses typical misreadings of corpus evidence, the scaffolding role of the teacher, and implications for corpus-assisted translation pedagogy in similar institutional contexts.