Research Article
BibTex RIS Cite

Yapay Zeka Kullanarak Dolandırıcı İnternet Sitelerinin Engellenmesi

Year 2025, Volume: 1 Issue: 1, 22 - 28, 30.06.2025

Abstract

Teknolojinin ve dijitalleşmenin hayatımızdaki yeri her geçen gün artarken, bu duruma ayak uyduran insanlar, şirketler ve devletler de siber saldırılara karşı daha savunmasız hale gelmektedir. Bu saldırılar arasında en yaygın olanı ise oltalama (phishing) saldırılarıdır. Bu saldırılarda dolandırıcılar, kimlik bilgilerinizi ve diğer hassas verilerinizi ele geçirmek için sahte web siteleri veya e-postalar kullanarak saldırıda bulunurlar. Siber güvenliğin öneminin her geçen gün artmasıyla birlikte, siber güvenlik şirketleri, akademisyenler ve devletler de bu tür saldırılara karşı oltalama önleme sistemleri geliştirmeye başlamıştır. Bu çalışmada, yapay zekâ alanında son yıllarda geliştirilen mimarilerin URL üzerinden oltalama saldırılarını tespit etme konusunda ne kadar etkili olabileceği araştırılmıştır. Araştırmada, farklı yapay zekâ mimarilerinin performansları karşılaştırılmıştır. Elde edilen sonuçlara göre, BERT mimarisi %98'lik doğruluk oranıyla en iyi performansı gösteren ağ olmuştur. DistilBERT mimarisi de yüksek test sonuçları verse de bazı URL'lerde hatalı sonuçlar vermiştir. CNN mimarisi ise Transformer mimarisine göre daha eski olmasına rağmen %91'lik başarı oranı elde etmeyi başarmıştır.

References

  • Chollet, F., & others. (n.d.). Keras. Retrieved May 6, 2024, from https://keras.io
  • Federal Bureau of Investigation, 2021. Internet Crime Report 2021. Retrieved May 10, 2024, from www.ic3.gov.tr
  • International Telecommunication Union., 2023. Statistics. Retrieved May 11, 2024, from https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx
  • Jain, A. K., & Gupta, B. B. ,2018. PHISH-SAFE: URL features-based phishing detection system using machine learning. In Advances in Intelligent Systems and Computing (Vol. 729, pp. 467–474). https://doi.org/10.1007/978-981-10-8536-9_44
  • Jawade, J. V., & Ghosh, S. N., 2021. Phishing Website Detection Using Fast.ai library. Proceedings - International Conference on Communication, Information and Computing Technology, ICCICT 2021. https://doi.org/10.1109/ICCICT50803.2021.9510059
  • Mittal, K., Gill, K. S., Chauhan, R., Singh, M., & Banerjee, D., 2023. Detection of Phishing Domain Using Logistic Regression Technique and Feature Extraction Using BERT Classification Model. 2023 3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023. https://doi.org/10.1109/SMARTGENCON60755.2023.10442975
  • PhishTank. (n.d.). What’s PhishTank. Retrieved May 4, 2024, from https://phishtank.org/faq.php#whatisphishtank.
  • Siddharth Kumar. (2019). Malicious And Benign URLs. Retrieved May 4, 2024, from https://www.kaggle.com/datasets/siddharthkumar25/malicious-and-benign-urls

Block Fraudulent Websites Using Artıfıcıal Intelligence

Year 2025, Volume: 1 Issue: 1, 22 - 28, 30.06.2025

Abstract

As technology and digitalization play an increasingly important role in our lives, individuals, businesses, and governments that adapt to this trend are also becoming more vulnerable to cyberattacks. Among these attacks, phishing attacks are the most common. In these attacks, scammers use fake websites or emails to obtain your login credentials and other sensitive information. With the growing importance of cybersecurity, cybersecurity companies, academics, and governments have also begun to develop anti-phishing systems to counter such attacks. This study investigates how effective the architectures developed in the field of artificial intelligence in recent years can be in detecting phishing attacks through URLs. The performance of different artificial intelligence architectures was compared in the study. According to the results, the BERT architecture was the best performing network with an accuracy rate of 98%. While the DistilBERT architecture also had high test results, it gave incorrect results for some URLs. The CNN architecture, on the other hand, achieved a success rate of 91%, although it is older than the Transformer architecture.

References

  • Chollet, F., & others. (n.d.). Keras. Retrieved May 6, 2024, from https://keras.io
  • Federal Bureau of Investigation, 2021. Internet Crime Report 2021. Retrieved May 10, 2024, from www.ic3.gov.tr
  • International Telecommunication Union., 2023. Statistics. Retrieved May 11, 2024, from https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx
  • Jain, A. K., & Gupta, B. B. ,2018. PHISH-SAFE: URL features-based phishing detection system using machine learning. In Advances in Intelligent Systems and Computing (Vol. 729, pp. 467–474). https://doi.org/10.1007/978-981-10-8536-9_44
  • Jawade, J. V., & Ghosh, S. N., 2021. Phishing Website Detection Using Fast.ai library. Proceedings - International Conference on Communication, Information and Computing Technology, ICCICT 2021. https://doi.org/10.1109/ICCICT50803.2021.9510059
  • Mittal, K., Gill, K. S., Chauhan, R., Singh, M., & Banerjee, D., 2023. Detection of Phishing Domain Using Logistic Regression Technique and Feature Extraction Using BERT Classification Model. 2023 3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023. https://doi.org/10.1109/SMARTGENCON60755.2023.10442975
  • PhishTank. (n.d.). What’s PhishTank. Retrieved May 4, 2024, from https://phishtank.org/faq.php#whatisphishtank.
  • Siddharth Kumar. (2019). Malicious And Benign URLs. Retrieved May 4, 2024, from https://www.kaggle.com/datasets/siddharthkumar25/malicious-and-benign-urls
There are 8 citations in total.

Details

Primary Language English
Subjects Machine Learning (Other), Software and Application Security, Artificial Intelligence (Other)
Journal Section Research Article
Authors

Semih Güner This is me 0009-0008-2969-1403

Büşra Takgil 0000-0002-7927-0083

Early Pub Date June 26, 2025
Publication Date June 30, 2025
Submission Date May 9, 2025
Acceptance Date June 10, 2025
Published in Issue Year 2025 Volume: 1 Issue: 1

Cite

APA Güner, S., & Takgil, B. (2025). Block Fraudulent Websites Using Artıfıcıal Intelligence. Siber Güvenlik Ve Dijital Ekonomi, 1(1), 22-28.