I searched online and i could not make sense of features used in finding if two questions are similar or not? So please help me explain the feature engineering used in detecting how two questions are duplicate or not? Also, let me know for a new question how can we detect the set of questions which are similar to each other?