Recognizing Depression from Twitter Activity

This paper extensively evaluates the effectiveness of using a user's social media activities for estimating degree of depression, and extracts several features from the activity histories of Twitter users to construct models for estimating the presence of active depression.