2026-05-19T14:09:46.167Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:10145 2025-01-03T14:40:07Z 3

Investigating the Use of Generative Adversarial Networks for Cybersecurity Dataset Training Data Substitution, Chapter in LNNS Lecture Notes in Networks and Systems: INFUS 2024: Intelligent and Fuzzy Systems, Springer, volume 1089 2024 10.1007/978-3-031-67195-1_81 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/10145 https://link.springer.com/chapter/10.1007/978-3-031-67195-1_81 M. Zivkovic A. Petrovic M. Sarac L. Jovanovic M. Antonijevic T. Zivkovic N. Bacanin Generative adversarial networks (GAN) in cybersecurity, particularly in natural language processing (NLP), remain underexplored. This study investigates the performance of machine learning (ML) techniques, specifically focusing on predicting spam emails. ML solutions offer customization and problem-specific tuning, making them well-suited for this task. The applied NLP approach employs term frequency-inverse document frequency (TF-IDF) for text mining, evaluating term importance based on overall appearance across documents and frequency in the evaluated document. The framework proves crucial for detecting patterns, especially in the absence of unfiltered spam email data. To address data scarcity, the study evaluates ML predictions on both real and GAN-generated data. The comparison reveals that the best models, trained on real and synthetic data, achieve over 97% accuracy, suggesting GANs as a promising approach to overcome limitations associated with limited secure data. bookPart Springer, Cham 741 749 LNNS Lecture Notes in Networks and Systems: INFUS 2024: Intelligent Industrial Informatics and Efficient Networks Proceedings of the INFUS 2024 Conference, volume 1089