SQL injection attack identification method based on deep learning

Abstract

The invention discloses an SQL injection attack recognition method based on deep learning, which comprises the following steps: S1, data preprocessing: carrying out data cleaning on related data, and manually constructing features according to data distribution; s2, word vector model generation: performing natural language processing on related data by using a neural network to fit a word vector model; s3, convolutional neural network training: generating a convolutional neural network for text content according to the word vector model, and performing feature extraction of related data; s4, Dropout layer training: using Dropout to reduce the interaction of the feature detection device; s5, related data full-connection network training: the local features of the related data after Dropout layer training are assembled into a complete graph again through the weight matrix; and S6, Softmax layer classification: carrying out Softmax function operation on the output of S5, and then judging whether related data is SQL injection attack or not. According to the method, the SQL injection attack recognition efficiency can be effectively improved, and the safety is improved.

Description

technical field [0001] The invention relates to the technical field of network data security, in particular to a method for identifying SQL injection attacks based on deep learning. Background technique [0002] Web-based applications typically rely on back-end database servers to manage application-specific persistent state and to extract data by executing queries composed using input provided by the application's users. If the user's request is not handled correctly, the attacker can change the structure of the SQL statement by manipulating user input. As a result, the server may suffer from SQL injection attacks based on web applications, and even endanger the security of the database. [0003] SQL injection attack detection is divided into pre-intrusion detection and post-intrusion detection. Pre-intrusion detection can be done manually or by using SQL injection tool software. The purpose of detection is to prevent SQL injection attacks, and the detection after SQL inj...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the technical problem that people cannot timely and effectively prevent SQL injection attacks when using Internet data in the prior art, and provides a method for identifying SQL injection attacks based on deep learning, which can effectively improve the efficiency of SQL injection attack identification and improve security sex

Images