Cognitive Cyber Crimes in the Era of Artificial Intelligence

19AI-Assisted Deep Social Profiling for Cyber Exploitation

Vivek Bhardwaj¹, Tanima Thakur², Jeyaganesh Viswanathan³ and Mukesh Kumar⁴*

¹Amity School of Engineering & Technology, Amity University Punjab, Mohali, Punjab, India

²School of Computer Science and Engineering, Lovely Professional University, Phagwara, India

³Zoetis, New Jersey, USA

⁴Advanced Centre of Research & Innovation (ACRI), Department of Computer Application, Chandigarh School of Business, Chandigarh Group of Colleges Jhanjeri, Mohali, Punjab, India

Abstract

This study presents a novel artificial intelligence–assisted deep social profiling framework aimed at detecting and mitigating cyber exploitation activities across social media platforms. Utilizing the Twitter Sentiment140 Dataset, which comprises 1.6 million labeled tweets, our approach leverages a hybrid graph neural network combined with transformer-based BERT (Bidirectional Encoder Representations from Transformers) embeddings to deeply analyze user behavior, network interactions, and content semantics. We introduce Social Exploitation Risk Index as a unique metric for profiling risk levels in user interactions. Experimental results demonstrate the model achieves an F1 score of 0.942, accuracy of 94.7%, and area under the curve of 0.961, outperforming baseline models by over 6%. This approach enables proactive identification of cyber exploitation vectors with enhanced precision and real-time adaptability.

Keywords: Deep social profiling, cyber exploitation, ...