International Journal of Computing and Mathematics

Considering that the quantity of emails globally is 269 billion messages per day and that 49.7% is spam that involves emails from fraudsters. These cybercriminals intend to "phish" their victims for private sensitive data or infect their pcs with viruses or malicious content for illegal economic gains. Therefore, this paper describes how these online scams are perpetuated and provides several inquiries and counterattack strategy suggestions by machine learning professionals to address the problem of spam filtering. This article reports various study designs and suggested solutions using machine learning algorithms, ranging from methods based on text categorization to approaches that examine email content with the attached pictures. The efficacy and efficiency of these machine learning instruments have been found and discussed. In conclusion, further study on spam filtering instruments based on machine learning algorithms has been urged as cybercriminals are continually innovating new techniques that threaten and abuse these technologies in their bid to prevent spam filters.