Using Machine Learning to Protect Users Accounts in Twitter

Twitter is one of the most used social media sites where millions of people interact daily. Users perform tremendous tasks on Twitter, and it also impacts their daily lives in several ways. As social media usage is higher, the higher is the risk of spammers and attackers on social media sites and the users. Privacy and security are the most crucial aspects that every user wants, and every social media site follows. Fake and spammers accounts are the most significant issues these days. Several fake accounts on Twitter can spread phony information, make fake user following, and create several other problems. There is a spread of malicious content laying on the names of several known celebrities, which began several controversies. Machine learning is the most used technology these days as it is mainly used in predicting and classifying data. This research uses machine learning technology to detect Twitter accounts to achieve high accuracy. A support vector machine algorithm is used in this regard. First, a Twitter dataset was uploaded to the notebook. A data preprocessing was then done on the Dataset, in which attributes were distributed or extracted, stop words removed, and all other raw information was removed from the Dataset. Then, by using sci-kit, learn first a model was trained by using preprocessed data. Then fake accounts were tested accordingly. A 96.4% accuracy was achieved by using a support vector machine algorithm.