TWEET BASED BOT DETECTION USING BIG DATA ANALYTICS

Abstract

Twitter, a leading micro-blogging platform with millions of users worldwide, has become a prime target for various malicious activities, including the dissemination of rumors, phishing attempts, and malware distribution. Among these threats, the proliferation of tweet-based botnets poses a significant risk, capable of orchestrating large-scale attacks and manipulative campaigns. To combat such threats effectively, the utilization of big data analytics techniques, particularly shallow and deep learning methodologies, has emerged as a viable solution for accurately distinguishing between human-operated accounts and bot-generated tweets. In this paper, we comprehensively review existing techniques and propose a taxonomy to categorize the state-of-the-art tweet-based bot detection methodologies. Furthermore, we delve into the intricacies of shallow and deep learning approaches employed for tweet-based bot detection, elucidating their respective performance outcomes. Lastly, we highlight the persistent challenges and unresolved issues within the domain of tweet-based bot detection, underscoring the importance of ongoing research efforts to mitigate emerging threats effectively.