Abstract
Http requests represent the main component of a web navigation system. These requests, once received by the server, need to be analyzed to guarantee that they are attack-free. Attacks carried by Http requests can have disastrous effects. Due to the importance of Http requests, it is crucial to design an efficient and robust Http request analyzer that guarantees the detection of malicious ones and prevent them from being processed. In this paper, we propose a new technique to process the Http request called Code Embedding. The proposed method was integrated within the Convolutional neural network to provide an efficient and robust web attack detection tool. The experimental results prove that our method outperforms the previous works and reaches an accuracy of 98.125%.