A Machine Learning Approach to Guard Social Media Accounts from Malicious Links

The number of Internet users worldwide is rapidly increasing, providing growth opportunities in business, education, sports, entertainment, and social networking. As more people connect to the Internet, security threats can become more dangerous. We have seen a rapid increase in the amount of information generated and shared on social media. The use of social media has increased exponentially, and the rapid growth of users is unprecedented in the history of the technological revolution. The continuous development of social media has created an ecosystem where users from different cultures and regions connect across platforms. However, the rapid growth of social media users has led to a high risk of information leakage and network security threats. Nowadays, hackers use spam methods containing malicious viruses to steal information from social media profiles. They send spam emails containing malicious links, and once the user verifies that they have clicked on those links, they can access the user’s profile information. In this article, we propose a spam filter using decision tree classifier and explain the different approaches to take care social media user’s profile. We measure the performance of different classifiers to find best model for emails_small and emails_full both datasets. We observed Decision Tree with information gain works more effectively than other classifiers.