FAKE PROFILE DETECTION ON SOCIAL NETWORKING WEBSITES

Abstract

In today’s digital age, the ever-increasing dependency on computer technology has left the average citizen vulnerable to crime such as data breaches and possible identity theft. These breaches or crime often target social media networks such as Instagram, Facebook, LinkedIn and Twitter. This emerges the needs for social networks to improve their cyber security .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 pro- file to be authentic/not. The parameters of the social media network 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 Face- book, 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, Random Forest and Neural Network models to determine fake profiles in social network with improved accuracy.