SQL Injection Attack Detection Using Machine Learning Techniques

Inter-processes shared information between applications that are running over multiple platforms over the Internet is the most appropriate environment to hack these data. Users emit a huge amount of confidential data daily via web applications that are stored in databases of these applications. That causes applications and their databases to be vulnerable to all kinds of information security threats. SQL injection attack is considered to be among the top ten vulnerabilities that threaten web applications. This type of attack allows the attacker to grab confidential information stored in databases and consequently harm the reputation of companies. This paper addresses SQL injection attacks and the related detection and identification methods. An intelligent method that relies on machine learning techniques and convolutional neural networks (CNN) is proposed to design a model that identifies those attacks and classifies SQL queries into malicious or healthy categories. The experimental results of the proposed model show a high classification accuracy that reaches 97.74% which outperforms the reported methods in the literature.