🤖 AI Summary
This study addresses the language barrier impeding knowledge exchange between Arabic- and Russian-speaking scientific communities, which hinders collaboration toward Sustainable Development Goals 9 and 17. To bridge this gap, the authors construct the first Arabic–Russian scientific parallel corpus, comprising approximately 27,000 sentence pairs, and perform efficient domain-specific fine-tuning of multilingual large language models—including mT5-base, NLLB, and Qwen2.5-7B-Instruct—using LoRA and QLoRA. Experimental results demonstrate that fine-tuned models significantly outperform few-shot prompting; notably, Qwen2.5-7B-Instruct with QLoRA (rank=8) achieves a BLEU score of 23.15 and a COMET score of 0.758. The project releases the corpus, fine-tuned models, and evaluation code, establishing the first benchmark for scientific translation between these two languages.
📝 Abstract
Russian and Arabic are among the major languages of scientific communication. Language barriers impede the exchange of research results between these communities, which affects international collaboration and the progress of sustainability-related research. We present a benchmark for Arabic--Russian scientific translation. The benchmark includes a hybrid parallel corpus of about 27,000 sentence pairs, compiled from scientific abstracts and general-domain texts (religion, news, conversations). We fine-tune three multilingual language models -- mT5-base (580M parameters), NLLB-200-distilled-1.3B (1.3B), and Qwen2.5-7B-Instruct (7B) -- using LoRA with ranks 8, 16, 32, and 64. The Qwen2.5-7B model with QLoRA (rank 8) yields BLEU 23.15, chrF 43.89, BERTScore 0.906, and COMET 0.758. These are +4.36 BLEU and +0.051 COMET above the zero-shot baseline. Few-shot prompting with three examples does not improve performance, indicating that domain-specific fine-tuning is required. We release the models, the corpus, and the evaluation code. By lowering the language barrier for scientific texts, the work enables knowledge exchange between Arabic-speaking and Russian-speaking researchers. It contributes to sustainable partnerships (UN SDG 17) and innovation infrastructure (SDG 9), aligning with the conference's focus on technology-driven sustainable development.