Bot Detection in Social Networks Based on Machine Learning Techniques, User Information and Activities

With the swift rise of social networking sites, they have now come to hold tremendous influence in the daily lives of millions around the globe. The value of one's social media profile and its reach has soared highly. This has invited the use of fake accounts, spammers and bots to spread content favourable to those who control them. Thus, in this project we propose using a machine learning approach to identify bots and distinguish them from genuine users. This is achieved by compiling activity and profile information of users on Twitter and subsequently using natural language processing and supervised machine learning to achieve the objective classification. Finally, we compare and analyse the efficiency and accuracy of different learning models in order to ascertain the best performing bot detection system.