Abstract
Phishing is a type of Social Engineering cyber-attack, hackers use it to gain access to confidential credentials like bank account credentials details, details of their personal life like debit card details, social media credentials, etc. Phishing website links seem to seem just like the genuine ones and it's a tedious and troublesome task to differentiate among those websites. In this paper, features are extracted from a separate dataset of phishing and benign website URLs and then using the Machine Learning method we determine the phishing websites. We also rank the features based on the contribution of each feature used in determining the outcome of a URL link using built python libraries. Most of the phishing URLs use a large URL length when used for an attack. Hence, we proposed three machine learning models Random Forest, Support Vector Machine (SVM), Decision trees models for the efficient detection of phishing using fake URLs. The performance of the models is also compared among themselves using a confusion matrix to determine the highest performance. The implemented models have shown an accuracy of 84.81 (for Random Forest and SVM),83.96 (Decision tree)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have