Manta Ray Foraging Optimizer with Deep Learning based Malicious Activity Detection for Privacy Protection in Social Networks

Malicious activity detection is a vital component of ensuring privacy protection in social media networks. As users engage in online interactions, protecting their sensitive information becomes paramount. Social networks can proactively identify and mitigate malicious behaviors, such as cyberbullying, data breaches, and phishing attacks by applying advanced AI and machine learning (ML) technologies. This detection system analyzes user behavior patterns, content, and network traffic to flag suspicious activities, thus safeguarding user privacy and fostering a safer online environment. The incorporation of robust malicious activity detection mechanisms helps maintain trust in social networks and reinforces the commitment to preserving user privacy in an increasingly interconnected digital landscape. This article introduces a novel Manta Ray Foraging Optimizer with Deep Learning based Malicious Activity Detection (MRFODLMAD) technique for privacy protection in social networks. The drive of the MRFODL-MAD technique is to detect and classify malicious activities in the social network. To accomplish this, the MRFODL-MAD technique preprocesses the input data. For malicious activity detection, the MRFODL-MAD technique employs long short term memory (LSTM) system. The MRFO algorithm has been executed to hyperparameter tuning process to improve the performance of the LSTM network. The experimental outcomes of the MRFODL-MAD algorithm can be tested on social networking database and the results inferred the improved performance of the MRFODL-MAD algorithm under various different measures.