Attention based Neural Machine Translation with Sequence to Sequence Learning on Low Resourced Indic Languages

Deep Learning techniques are powerful in mimicking humans in a particular set of problems. They have achieved a remarkable performance in complex learning tasks. Deep learning-inspired Neural Machine Translation (NMT) is a proficient technique that outperforms traditional machine translation tasks. Performing machine-aided translation on Indic languages has always been a challenging task considering their rich and diverse grammar. The neural machine translation has shown quality results compared to the traditional machine translation approaches. The task becomes problematic when it comes to low-resourced language like Sanskrit. This paper has presented our work on developing a dataset with Sanskrit as a language pair and an attention mechanism-based neural machine translation model trained on the Sanskrit-Hindi bilingual dataset. We have shown the significance of the attention mechanism to overcome the problem of long-term dependencies. The attention mechanism has shown promising results on low-resourced Indic languages.