Are Large Language Models State-of-the-art Quality Estimators for Machine Translation of User-generated Content?
Shenbin Qian, Constantin Orǎsan, Diptesh Kanojia, Félix do Carmo · 2024
This paper investigates whether large language models (LLMs) are state-of-the-art quality estimators for machine translation of usergenerated content (UGC) that contains emotional expressions, without the use of reference translations.To achieve this, we employ an existing emotion-related dataset with humanannotated errors and calculate quality evaluation scores based on the Multi-dimensional Quality Metrics.We compare the accuracy of several LLMs with that of our fine-tuned baseline models, under in-context learning and parameter-efficient fine-tuning (PEFT) scenarios.We find that PEFT of LLMs leads to better performance in score prediction with human interpretable explanations than fine-tuned models.However, a manual analysis of LLM outputs reveals that they still have problems such as refusal to reply to a prompt and unstable output while evaluating machine translation of UGC.