How Good is Zero-Shot MT Evaluation for Low Resource Indian Languages?
Indian Institute of Technology, Madras · National Institute of Information and Communications Technology (NICT), National Institute of Advanced Industrial Science and Technology · A*STAR · Microsoft and Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology · Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。DOI 10.18653/v1/2024.acl-short.58 ↗
摘要
While machine translation evaluation has been studied primarily for high-resource languages, there has been a recent interest in evaluation for low-resource languages due to the increasing availability of data and models. In this paper, we focus on a zero-shot evaluation setting focusing on low-resource Indian languages, namely Assamese, Kannada, Maithili, and Punjabi. We collect sufficient Multi-Dimensional Quality Metrics (MQM) and Direct Assessment (DA) annotations to create test sets and meta-evaluate a plethora of automatic evaluation metrics. We observe that even for learned metrics, which are known to exhibit zero-shot performance, the Kendall Tau and Pearson correlations with human annotations are only as high as 0.32 and 0.45. Synthetic data approaches show mixed results and overall do not help close the gap by much for these languages. This indicates that there is still a long way to go for low-resource evaluation.