Behavioural Feature Engineering for Phishing Detection: The Case of Redirection Patterns

Phishing detection remains a persistent challenge due to the continually evolving tactics of attackers. This study enhances phishing detection by emphasising redirection pattern features as key behavioural indicators. A two-phased feature extraction approach was employed: Phase 1 prioritised redirection patterns, such as the number of redirections and intermediate domain reputation, while Phase 2 incorporated supporting lexical and domain-based attributes. Using publicly available datasets from PhishTank and OpenPhish, the system was evaluated through multiple machine learning models. Results highlight the critical role of redirection patterns in capturing dynamic behaviours often overlooked by traditional feature sets. The modular feature extraction process not only improves detection accuracy but also offers adaptability to emerging phishing tactics. These findings provide a foundation for integrating behavioural insights into phishing detection frameworks and advancing the development of more resilient cybersecurity systems. Keywords: Behavioural Features, Feature Extraction, Phishing Detection, Phishing Tactics, Redirection Patterns, URL Analysis. Journal Reference Format: Tajudeen Mashkur Muhammad, Agwom Itserim Emmanuel, Fatimah Adamu-Fika, Kamal Shehu Bature, Auwal Bello, Onyinye Vivian Okpoko, & Aanuoluwapo Enyojo Baba-Onoja (2025): Behavioural Feature Engineering for Phishing Detection: The Case of Redirection Patterns. Journal of Behavioural Informatics, Digital Humanities and Development Res. Vol. 11 No. 2. Pp 106-114. https://www.isteams.net/behavioralinformaticsjournal.dx.doi.org/10.22624/AIMS/BHI/V11N2P210