A Systematic Review on Social Bots Account Detection Using Machine Learning
Pankaj Kumar Sharma, Tapsi Nagpal, Gulshan Shrivastava, Javalkar Dinesh Kumar · 2023
Online social networking sites are becoming more important in today’s digital age. Cybercriminals are interested in developing dangerous bots due to the large quantity of information available and their open nature. In these systems, malicious bots are computer programs or botnets designed to do harm by imitating human users. Moreover, such bots raise significant public opinion and cyber security issues. They are used for nefarious purposes including spamming, creating phony profiles, disseminating offensive or inaccurate information, engaging in click farming, stealing trending hashtags, and much more. Researchers and cyber criminals are always at odds, with researchers constantly developing new and improved bots to counteract the most up-to-date forms of detection technology. This study examines the most up-to-date breakthroughs in Machine Learning-based algorithms for bot detection and categorization across Facebook, Instagram, LinkedIn, Twitter, and Weibo, the five most prominent social media platforms. Herein, we provide a condensed overview of the many supervised, semi-supervised, and unsupervised methods, as well as details on the datasets used by the researchers. We also give a detailed classification of the features that were extracted. Also, this paper provides a quick summary of the difficulties and gains made in this field, as well as possible future research avenues and intriguing areas to investigate.