A deep learning classifier that predicts textual entailment and semantic relatedness between English sentence pairs (SemEval-2014 Task 1).

We achieved 82% test accuracy on the entailment subtask and 0.83 Pearson correlation score on the relatedness subtask, both of which comparable to the top three participants of the shared task at SemEval-2014.

• Assignment 1: Text pre-processing techniques
• Assignment 2: Training a deep learning network