Translate, then Detect: Leveraging Machine Translation for Cross-Lingual Toxicity Classification
This work addresses the problem of scalable content moderation for practitioners by providing actionable guidance on cross-lingual toxicity classification, though it is incremental as it builds on existing translate-test paradigms.
The paper tackled the challenge of multilingual toxicity detection by comparing translation-based and language-specific classification pipelines, finding that translation-based methods outperformed out-of-distribution classifiers in 81.3% of cases (13 of 16 languages) and that traditional classifiers beat large language model judges, especially for low-resource languages.
Multilingual toxicity detection remains a significant challenge due to the scarcity of training data and resources for many languages. While prior work has leveraged the translate-test paradigm to support cross-lingual transfer across a range of classification tasks, the utility of translation in supporting toxicity detection at scale remains unclear. In this work, we conduct a comprehensive comparison of translation-based and language-specific/multilingual classification pipelines. We find that translation-based pipelines consistently outperform out-of-distribution classifiers in 81.3% of cases (13 of 16 languages), with translation benefits strongly correlated with both the resource level of the target language and the quality of the machine translation (MT) system. Our analysis reveals that traditional classifiers outperform large language model (LLM) judges, with this advantage being particularly pronounced for low-resource languages, where translate-classify methods dominate translate-judge approaches in 6 out of 7 cases. We additionally show that MT-specific fine-tuning on LLMs yields lower refusal rates compared to standard instruction-tuned models, but it can negatively impact toxicity detection accuracy for low-resource languages. These findings offer actionable guidance for practitioners developing scalable multilingual content moderation systems.