Fake Profile Detection Using Machine Learning

Abstract

Platforms for social media like Facebook, Twitter, Instagram, and others have a big impact on our lives. All across the world, people are actively engaged in it. But, it also needs to address the problem of false profiles. Fake accounts are regularly made by people, software, or machines. They are employed in the spread of rumors and illegal actions like phishing and identity theft. This project uses several machine learning techniques to discriminate between fake and authentic Twitter profiles based on characteristics such as follower and friend counts, status changes, and more. Twitter profile dataset, classifying genuine accounts as TFP and E13 and fake accounts as INT, TWT, and FSF. In this section, the author talks about neural networks, LSTM, XG Boost, and Random Forest. The important traits are picked to judge the veracity of a social media page. The architecture and hyperparameters are also discussed. Lastly, after the models have been trained, results are generated. As a result, the output is 0 for true profiles and 1 for fake profiles. It is possible to disable or delete a fake profile when it is found, preventing cyber security issues.