Fake-News-Detection

Largely an unsupervised learning problem, since there are very few labelled datasets for supervised learning approach. The unsupervised approach largely takes the problem as an anomaly detection task wherein documents deviating from the general character are labelled as fake, inculcating a general assumption that most of the news are not fake.