Review and Evaluation of Classification Algorithms Enhancing Internet Security

This paper explores online learning and batch algorithms for detecting malicious Web sites (those involved in criminal scams) using lexical and host-based features of the associated URLs. A data set has been built including malicious and benign URLs, and data mining system Weka has been used as an aid to classify the existent URLs and new coming URLs and evaluate the classification algorithms. A real-time malicious URL detection system has been constructed successfully. The experiment result shows that this method can help to reduce internet access risk effectively.