Code-mixed and Low Resource Language Processing

Research > Code-mixed and Low Resource Language Processing > Code-Mixed Machine Translation

Code-Mixed Machine Translation

Overview

Humans have the unique ability to seamlessly switch between multiple languages in a single sentence. However, the present SOTA translation systems are incapable of translating these mixed-sentences. We intend to propose Neural Machine Translation (NMT) models which can accurately interpret and provide meaningful translations of code-mix sentences.