Phishing acquires sensitive information, like login credentials, i.e. usernames, passwords or other security tokens, card information, etc from users. The proliferation of which is a threat to our computing system and its’s security [1]. In most cases, Phishing does not require high technicality, making it the third most performed attack according to multiple sources. Despite many solutions being proposed to completely eradicate or at least mitigate about 80% of Phishing attacks, Phishing continues to be prevalent. Vishakha P.R. and Sahil S.J. [2], were unable to detect web pages automatically, and the systems were limited to system applications alone. Therefore, this work aims to design a system for detecting phishing attacks, Implement the design and evaluate the algorithm’s performance. Jupyter Notebook was the medium used for analysis. The two algorithms used were two-class logistic detection and Naïve Bayes. The dataset (phishing.csv) obtained from Kaggle was divided into two datasets to train and test in the ratio 80:20. The result of two class logistic regression was 94.2% compared to 85% test accuracy by the Naïve Bayes algorithm.

Authors: Olaniyi A. Ayeni, Hope O. Akinyemi

Published in: International Conference on Information Society (i-Society-2024)

• Date of Conference: 26-28 August, 2024
• DOI: 10.20533/iSociety.2024.0009
• ISBN: 978-1-913572-72-3
• Conference Location: Churchill College, Cambridge, UK