Fake Account Detection using Artificial Intelligence and Neural Networks

Abstract

Social media platform is extremely popular these days and because of it's open nature it leads to increase in cybercrimes, harassments, spam messages through fake accounts so the project focuses on detecting fake accounts on social media using SVM classifier(Support Vector Machine) and Neural Network model. Using a new algorithm that is SVM-NN. Here we are using MIB dataset which is a predefined dataset that consists of a list of features available in social media like the friend-count, request accept count, etc. Like that 16 top categories are the application for everything social media platform. Detecting which activity to perform fake account detection we can't predefine as it will be different for different accounts and social media platforms this issue is getting solved by SVM by selecting which of these activities given by the user that we have taken to identify which account is fake and which is not.