Small language models improve multilingual text classification via selective translation

Researchers have developed a cost-efficient routing pipeline for multilingual short-text classification that leverages small language models. This pipeline selectively translates lower-resource languages into English before classification, aiming to improve performance without requiring task-specific fine-tuning. Evaluations on two benchmarks, SIB-200 and MASSIVE, demonstrated that this selective translation approach significantly boosts the Macro-F1 scores for weaker languages, though the optimal strategy varies by task. AI

IMPACT Offers a more efficient approach to multilingual text classification, potentially improving operational systems like content moderation and customer support routing.

RANK_REASON Academic paper detailing a new methodology for language model application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Hugging Face
• MASSIVE

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →