## External dataset:

We need two types of datasets one with tweets containing depressive characteristic which is obtained from twitter API and the other one with  random tweets which is available in one of the [Kaggle datasets](https://www.kaggle.com/ywang311/twitter-sentiment/data).