AUTOMATED DETECTION OF POTENTIALLY DANGEROUS URL ADDRESSES USING THE SCIKIT-LEARN LIBRARY

Abstract

The methodology of automated detection of potentially dangerous URLs using the sci-kit-learn library is considered. The proposed methodology includes data preparation, feature generation, and model evaluation based on the random forest algorithm for classifying URLs into phishing and safe ones. The methodology is implemented using the Python programming language and the scikit-learn library. Experimental results show the effectiveness of the model in identifying potentially dangerous URLs, which plays an essential role in protecting users from fraud and other online threats.