The world’s first dataset aimed at improving the quality of English-to-Malayalam machine translation – a long-overlooked language spoken by more than 38 million people in India â€“ has been developed by ...