SQLIML: A Comprehensive Analysis for SQL Injection Detection Using Multiple Supervised and Unsupervised Learning Schemes

SQLIML: A Comprehensive Analysis for SQL Injection Detection Using Multiple Supervised and Unsupervised Learning Schemes

연구 분야: Databases

학회: SN Computer Science

SQL injection is a security concern that affects the majority of program’s that are built upon various types of databases. According to researchers, it is found that SQL injection is mainly vulnerable to database-driven software. The disadvantages of previously developed SQLI models include their inability to identify novel patterns; instead, they will only detect those that they have previously seen or have been trained on. Our model will be capable of determining how much of the data being entered has been SQL-injected by looking for patterns in the input. Our system will be able to recognize all injection procedures. All attributes will be extracted and chosen by the model. The text should only have to be entered once by the user. It can also be scaled and adapted to a variety of purposes.