Authors: Monther Aldwairi, Rami Alsalman
Surfing the World Wide Web is becoming a dangerous everyday task with the WWW becoming rich in all sorts of attacks. Websites are a major source of many scams, phishing attacks, identity theft, SPAM commerce and malwares. We propose a system to detect malicious websites based on URL and host features and call it MALURLs. The system uses Naïve Bayes classifier as a probabilistic model to detect if the target website is a malicious or benign. It introduces new features and employs self learning using Genetic Algorithm to improve the classification precision. A small dataset is collected and expanded through GA mutations to learn the system over short time and achieve an average precision of 89%.
Keywords: security; malicious websites; machine learning; genetic algorithm; classification.