– In today’s worlds where people are more reliable on the news which are available online as it's convenient for them. Fake news is one of the biggest new-age problems has the potential to mild opinions and influence decisions. The proliferation of fake news on social media and the Internet is deceiving people to an extent that needs to be stopped. The existing systems are inefficient in giving a precise statistical rating for any given news claim. Also, the restrictions on input and category of news make it less varied. This paper proposes a system that classifies unreliable news into different categories after computing an F-score. In this system, we’ve used Logistic Regression to classify fake news. The pre-processing functions perform some operations like tokenizing, n-grams, and exploratory data analysis. Simple Count Vectorization, TF-IDF is used as feature extraction techniques. The logistic regression and Multinomial model are used as a classifier for fake news detection with a probability of truth.