Twitter BOT Detection using Machine Learning

In todays world many people like a businessmans, Media, politicians, etc., uses Twitter daily & have become an important part of life. An increasing number of people on twitter but hide their identity for malignant purpose. It is dangerous for other users hence the necessity for identifying the twitter bots. Thus there is a developing need for distinguishing which account contains bots or not. The characteristics of twitter accounts are utilized as Features in machine learning algorithms to label users as genuine or fake. In this paper, we used three machine learning algorithms to detect the account is fake or real, which are Decision Tree, Random Forest, and Multinomial Naive Bayes The classification performance of the algorithms is compared with their accuracy. The accuracy given by the Decision tree algorithm is 93%, the Random Forest algorithm is 90% and the Multinomial Naive Bayes is 89%. Hence it is seen that the Decision tree gives more accuracy as compared to Random Forest and Multinomial Naive Bayes.