Complex snort rule classification method and system based on depth features

The invention provides a complex snort rule classification method and system based on depth features. The method comprises the steps that according to snort rules, a multi-stage classifier based on content depth features, a weak classifier based on a sparse matrix and a correlation classifier are built respectively, wherein the multi-stage classifier based on the content depth features is built according to content keyboards in the snort rules, the weak classifier based on the sparse matrix is built according to combination relevant keywords in the snort rules, and the correlation classifier is built according to the preposition rule and the postposition rule in the snort rules; in semi-supervised learning, the multi-stage classifier based on the content depth features, the weak classifier based on the sparse matrix, the correlation classifier and a one-dimensional SVM classifier in the snort rules are trained, a complex snort rule total classifier based on the depth features is formed, and therefore classifi