Threat Detection Using Internet Agents and Text Categorization

摘要

Previous work in intrusion detection has been focused on techniques internal to networks and hosts, which are insular and lack pro-active capabilities. Relatively little research has been done in the area of detecting threats externally, using information gathered on the Internet. We introduce threat detection through the integration of text categorization and Internet Agent technologies. We create an intelligent agent, Mewshound that is in use by law enforcement, and can be trained to detect categories related to cyber-terrorism. The categories are focused on the threat.