SSA-COMET: Do LLMs Outperform Learned Metrics in Evaluating MT for Under-Resourced African Languages?

TL;DR

This paper introduces SSA-MTE, a large-scale human-annotated dataset for African language MT evaluation, and develops SSA-COMET metrics, demonstrating they outperform existing models and are competitive with top LLMs, especially for low-resource languages.

Contribution

The paper presents SSA-MTE dataset and SSA-COMET metrics, advancing evaluation of MT for African languages and benchmarking LLMs in this context.

Findings

SSA-COMET outperforms AfriCOMET in evaluations.

SSA-COMET is competitive with GPT-4o, Claude-3.7, and Gemini 2.5 Pro.

Resources are openly available for future research.

Abstract

Evaluating machine translation (MT) quality for under-resourced African languages remains a significant challenge, as existing metrics often suffer from limited language coverage and poor performance in low-resource settings. While recent efforts, such as AfriCOMET, have addressed some of the issues, they are still constrained by small evaluation sets, a lack of publicly available training data tailored to African languages, and inconsistent performance in extremely low-resource scenarios. In this work, we introduce SSA-MTE, a large-scale human-annotated MT evaluation (MTE) dataset covering 14 African language pairs from the News domain, with over 73,000 sentence-level annotations from a diverse set of MT systems. Based on this data, we develop SSA-COMET and SSA-COMET-QE, improved reference-based and reference-free evaluation metrics. We also benchmark prompting-based approaches using state-of-the-art LLMs like GPT-4o, Claude-3.7 and Gemini 2.5 Pro. Our experimental results show that SSA-COMET models significantly outperform AfriCOMET and are competitive with the strongest LLM Gemini 2.5 Pro evaluated in our study, particularly on low-resource languages such as Twi, Luo, and Yoruba. All resources are released under open licenses to support future research.