DGA domain name family multi-classification detection method based on RBL-CNN-MA

The invention provides a DGA domain name family multi-classification detection method based on RBL-CNN-MA. The method comprises the following steps: inputting to-be-classified DAG domain name data into a RoBERTa pre-training model to obtain a sentence embedding feature vector and a word embedding feature vector of a domain name; inputting the character embedding feature vector into a bidirectional LSTM layer to extract long-distance features, and obtaining a text matrix fusing domain name context relevance; inputting the text matrix into a CNN layer to extract local features; inputting the output of the CNN layer to a multi-head attention layer to perform feature extraction from multiple dimensions; fusing the output of the multi-head self-attention layer with the sentence embedding feature vector output by the RoBERTa pre-training model to obtain a new sentence embedding feature vector; and inputting the new sentence embedded feature vector into a full connection layer to obtain a classification result of th