• Prof. K.S. Sawant
• Sejal Bora
• Shravani Thakare
• Mahak Chawla
• Prajakta Khairnar

Phishing attacks involve creating replica web pages to trick users into revealing personal information. This paper introduces an end-host based anti-phishing algorithm called SVM, which utilizes the generic characteristics of hyperlinks in phishing attacks. Derived from the Anti-Phishing Working Group's data, SVM can detect both known and unknown phishing attacks. Implemented on Windows, SVM successfully detected 195 out of 203 phishing attacks in experiments, demonstrating minimal false negatives, lightweight performance, and real-time detection capabilities. Additionally, using machine learning algorithms to predict malware infection rates, Light GBM emerged as the best model with an AUC Score of 0.73926, based on a publicly available dataset.

- SVM, Known and unknown phishing detection, Machine learning algorithm, malware infection rates, phishing, hyperlink characteristics, windows platform, Realtime detection, light GBM, public safety.

"Identification of security threats and legitimate status", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 6, page no.h129-h132, June-2024, Available :http://www.jetir.org/papers/JETIR2406712.pdf

"Identification of security threats and legitimate status", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 6, page no. pph129-h132, June-2024, Available at : http://www.jetir.org/papers/JETIR2406712.pdf