Identifikasi Akun Palsu Instagram Menggunakan Support Vector Machine

Abstract

One of the social media sites that has the most active users is Instagram, which is in fourth place compared to other social media sites. At this time, not a few people who use Instagram are competing to increase the number of followers in order to gain exposure or the trust of the wider community. Therefore, quite a few Instagram users create fake accounts to increase their followers. Apart from that, these fake accounts are also used to commit crimes such as fraud and cyberbullying. With so many fake accounts popping up, a system is needed to identify Instagram accounts, whether they are fake accounts or real accounts, in order to provide a sense of comfort and security for Instagram users and reduce crime on Instagram. In this study, the identification of accounts on Instagram social media will be carried out to determine the authenticity of the account by creating a website-based system. Identification was carried out using the Support Vector Machine method with a Radial Basis Function (RBF) kernel. The identification process begins with data extraction, then data labeling, until the classification process uses a support vector machine. The evaluation method used in this research is a confusion matrix, which manifests an accuracy value of 94%. After conducting several evaluations, it was concluded that the algorithm used by researchers can identify real and fake accounts on Instagram with good performance.