A Study on AI Models for Korean Translation Evaluation : Focusing on Multi-layered Automatic Evaluation Metrics (AEMs)

Using Multi-layered Automatic Evaluation Metrics (AEMs) like AEM-L2, AEM-S1, and AEM-M1, this study develops a model to evaluate translation quality across form, syntax, and semantics. Its implementation offers valuable feedback for learners and professionals. It also suggests integrating Speech-to-Text and Text-to-Speech technologies to assess interpretation accuracy and naturalness. Future research will focus on more diverse, pragmatic evaluation metrics. (Hankuk University of Foreign Studies)