This Repository consists files of the NLP Project - Fake News Detection Classifier which was held as a Data Science assessment by Techigai ,Hyd.
Fake News Detection
Problem Statement : It has become extremely easy to propagate any information to the masses, thus making Fake News a potential threat to public discourse. We would like you to build a model that when given a news article should determine the relevance of the body and the claim and classify the article accordingly into four categories: ‘agree’, ‘disagree’, ‘discuss’ and ‘unrelated’. Classes ‘disagree’ and ‘unrelated’ provide strong evidence that the news is fake, ‘agree’ indicates that the news is genuine.
Validation Metric F1-Score