FAKE PROFILE DETECTION ON SOCIAL NETWORKING WEBSITES

Abstract— The increased reliance on computer technology in today’s digital age has made the common citizen exposed to crimes such as data breaches and probable identity theft. These breaches or crime often target social media networks such as Instagram, Facebook, LinkedIn and Twitter. The motivates the needs for social networks to improve their cyber safety. The project intends to build an artificial intelligence solution to prevent the danger of a bot in the form of a fake profile on social media .The deep learning algorithms can determine the possibility of a social media profile to be authentic/not. The attributes of the social media that drive a breach are also identified in the work and web browser plugin is built to identify these fake profiles . Researches have observed that 20 percentage to to 40 percentage profiles in online social networks like Facebook, Instagram, linkedIn and Twitter are fake profiles. The owners of fake accounts extract the personal information about other people and spread the forget data on social networks. The Work employs a combination of SVM and Random Forest model and Neural Network model to determine fake profiles in social network with improved accuracy.