Rumor Detection

Abstract

Everyone has internet access and is connected to social media in today's fast-paced world. Numerous pieces of data are disseminated on these websites, but there is no reliable source for confirmation or verification. This is where rumors come into play. Rumors are deliberate fabrications intended to sway or drastically alter popular opinion, and their impact can be seen in politics, especially during elections, and on social media. Thus, to resolve this problem, a rumor detector is needed that is capable of accurately indicating whether information is false or real. We implemented algorithms such as Multinomial Naive Bayes, Gradient Boosting, and Random Forest on complex datasets to get this Rumor Detection System closer to more reliable rumor performance. Accuracy of Multinomial Naive Bayes is approximately 90.4Forestitwas86.588.3