With the advent of the World Wide Web and the swift advocacy of online platforms paved the way for news propagation that has never been seen in the past. With the present situation of social media platforms, users are developing and sharing more information when compared to the last five years, some of them are not even related to real life. Classifying the text automatically is a tedious and tough job to do. To give a verdict on the truthfulness of an article, a professional too needs to explore multiple aspects of the domain first. Machine learning algorithms are popularly being used to detect the truthfulness of a piece of text. In present scenario, different performance metrics are used to compare and evaluate the effectiveness of various machine learning algorithms. The study examines various textual properties that can be utilized to differentiate the fake and real news. Natural Language Processing techniques are used for data pre-processing which increases the accuracy of the machine learning models. Further, the extracted and preprocessed properties are used to train various ML classifiers with all possible combinations and the built models are then evaluated using various performance metrics.