A systematic literature review on spam content detection and classification

Area of Science:

• Computer Science
• Information Security
• Social Media Analysis

Background:

• Social media usage has surged, increasing exposure to spam content like malicious links and fake news.
• Users struggle to identify spam messages, posing risks to personal information and online safety.
• Effective spam detection is vital for maintaining social media security and user trust.

Purpose of the Study:

• To provide a comprehensive survey of recent advancements in spam text detection and classification on social media platforms.
• To analyze various techniques, including Machine Learning, Deep Learning, and text-based approaches, for combating social media spam.
• To highlight challenges in spam identification and discuss control mechanisms and relevant datasets.

Main Methods:

• Literature review of existing research on spam detection in social media.
• Categorization and discussion of Machine Learning algorithms applied to spam classification.
• Analysis of Deep Learning models and traditional text-based methods for spam identification.

Main Results:

• Identified a growing trend in spam content across platforms like Facebook, Twitter, and YouTube.
• Detailed the application of various AI techniques for effective spam text detection and classification.
• Summarized key challenges, control strategies, and datasets utilized in current spam detection research.

Conclusions:

• Spam detection and classification are essential for social media security.
• Machine Learning and Deep Learning offer promising solutions for identifying and mitigating spam.
• Further research is needed to address the evolving challenges in spam control mechanisms.