A framework for Twitter spam detection and reporting

Twitter is an American social networking and blogging site where users can communicate through messages known as "Tweets. The number of characters is restricted to 280 for all languages except Korean, Chinese and Japanese. Users on social media platforms tend to easily believe the contents of posts related to any random events and some of these events happen to be fake. Twitter spammers may fulfill their malicious objectives, such as spam sending, distributing malware, hosting botnet command and control (C&C) networks, and launching other illicit activities in the underground. Hence, we have proposed a system that will not only help in finding the type of spammers but also eliminate identical tweets. So we have implemented multi-classifier algorithms such as naive Bayes, K-Nearest neighbor, Decision tree and Random forest on a dataset obtained from Twitter and their performance is compared, also the most accurate algorithm is found out. The results of the experiment have been very positive.