Research on malicious links detection system based on script text analysis

Massive trojan and virus spreading via web vulnerabilities has become a severe threat to today´s cyber-society. To traditional honeypots, which is able to collect malware with automated manner, becomes one of the most severe and common detection methods nowadays. Compared with honeypots, malicious links detection system based on script text analysis has several advantages due to its lower cost and easier deployment. Firstly, according to the current analysis of trojan web links and existing problems, we summarize the requirement for malicious links detection system, and describe the system design based on requirement. After that we describe the algorithm with mathematical meathod. Finally, we analyze and summarize the experimental results, and verify the reliability and rationality of malicious links by this detection system.