HTTP Botnet Detection Algorithm Based on Content Association Recommendation

摘要

HTTP botnet is widely distributed and causes great harm. The traditional detection method is not analyze the attack data stream until the attack stop. In order to further reduce the harm, according to the characteristics of http/https botnet, an online detection method based on HTTP protocol is proposed, which is based on the content association recommendation algorithm. This method is able to distinguish between normal data and malicious data, and complete detection before the attack start, without increasing the burden of the network because of small complexity. The experiment proves the feasibility of this method.