A Novel 3 Stage Hybrid Framework For SQLIA Detection, Prevention Using Pattern Based And Supervised Machine Learning Model

Computer Engineering

Internet plays a very important role in our day to day life. Web application provides various services. As usage of web application increased, security become major concern. Security experts from SANS, a major SQL injection threat affected 160000 websites which are using IIS , Apache and MYSQL frameworks due to lack of proper validation and security holes SQL injection. This paper introduces two level framework for detection and prevention of SQLIA using patterns based and machine learning models. Despite of the abundance of techniques available for preventing SQL injection, issues, recently vulnerability database reported new major SQL injection threats. So it is necessary. In proposed framework query will pass through two stages scanning for SQLIA detection and prevention. The two stages are pattern based and machine learning based. As well database firewall will be there for monitoring incoming SQL queries. If query passes all this stages, then only query is valid otherwise there is attack. The approach usage a patterns and machine learning models in which value is entered for every field is checked for SQL injection attack by parsing it.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.