Somethingphishy: A Phishing Detection Browser Extension Using Logistic Regression
Jerian R. Peren, Mhike Jolo Y. Cron, Mark Simon S. Fabros, Danah Janyll M. Amoroso · 2025
The increased amount of phishing attacks during the COVID-19 Pandemic and subsequent years has prompted worries in the cybersecurity sector. Despite technological improvements and the implementation of new regulations and rules in the Philippines, phishing attacks remain a widespread and serious danger. The purpose of this investigation is to create an automated phishing detection browser plugin for Firefox and Chromium browsers. The plugin will be capable of identifying phishing links by analyzing various URL features, including the frequency of each special character and the number of redirects. Apart from other functions like allow and block lists and an appeals page for website owners, the mentioned browser plugin would be able to predict phishing websites and prohibit them. Tested on 1802 links, the plugin displays accuracy of 83.24 %, precision of 77.04 %, recall of 85.77 %, F1 score of 81.17 %, and ROC-AUC of 92.55 %. Following a “Acceptable” comment from the software quality evaluation questionnaire based on the ISO/IEC 25010:2023 criterion, the program also responded favorably.