Abstract :

The Rapid Growth Of Anonymous Online Platforms Has Enabled Cybercriminals To Engage In Illicit Activities Ranging From Drug Trafficking And Weapon Trade To Malware Distribution And Human Exploitation [7][10][12]. Detecting And Classifying Such Malicious Content On The Dark Web Remains Extremely Difficult Due To Its Unstructured Textual Nature, Hidden Linguistic Patterns, And Constantly Evolving Terminology [1][4][15]. To Address These Challenges, This Research Proposes A Hybrid Deep Learning Framework That Integrates Text-Convolutional Neural Networks (Text-CNN) With Topic Weight Embedding For Intelligent Dark Web Content Classification [3][5][8]. The Approach First Extracts Latent Semantic Patterns Using Topic Modeling, Where Topic Probability Weights Are Incorporated Into The CNN Input Layer To Enhance Contextual Awareness During Feature Learning [2][6][15]. The Combined Architecture Captures Both Local N-gram Dependencies And High-level Thematic Relevance, Delivering A Richer Representation Of Criminal Intent And Activity Indicators [5][8][9]. Experiments Conducted On Multilingual Dark Web Forum Datasets Demonstrate That The Proposed Model Significantly Outperforms Conventional Machine Learning, Pure CNN, And Standalone Topic Modeling Methods In Terms Of Accuracy, Precision, Recall, And F1-score [1][3][9][11]. Furthermore, The System Provides Interpretable Topic-to-category Associations, Enabling Security Researchers And Law-enforcement Analysts To Better Understand Emerging Criminal Trends [11][13][19]. This Hybrid Text-CNN With Topic Weight Embedding Constitutes A Robust And Scalable Solution For Proactive Cyber-crime Intelligence, Improving Automated Surveillance And Threat Monitoring Across Hidden Digital Ecosystems [8][13][20]. Keywords : Dark Web Monitoring, Text-CNN, Topic Modeling, Topic Weight Embedding, Deep Learning, Cybercrime Detection, Illicit Content Classification, Threat Intelligence.

Published:

29-11-2025

Issue:

Vol. 25 No. 11 (2025)

Page Nos:

301-307

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite