Unified NMT models for the Indian subcontinent transcending script-barriers

Anonymous

Abstract: Highly accurate machine translation systems are very important in societies and countries where multilinguality is very common, and where English often does not suffice. The Indian subcontinent is such a region, with all the Indic languages currently being under-represented in the NLP ecosystem. It is essential to advance the state-of-the-art of such low-resource languages atleast by using whatever data is available in open-source, which itself is something not very explored in the Indic ecosystem. In our work, we focus on improving the performance of very-low-resource Indic languages, especially of countries in addition to India. Specifically, we propose how unified models can be built that can exploit the data from comparatively resource-rich languages of the same region. We propose strategies to unify different types of unexplored scripts, especially Perso-Arabic scripts and Indic scripts to build multilingual models for all the Indic languages despite the script barrier. We also study how augmentation techniques like back-translation can be made use-of to build unified models that achieve state-of-the-art result among open source models, especially just using openly available raw data.