Project classifies Hinglish social content on platforms like Twitter, Reddit.
problem Classifying abusive and hate-inducing content in Hinglish on social media.
method Used deep learning with bi-directional sequence models and text augmentation techniques.
result Produced a state-of-the-art classifier that outperforms previous work.