Research Article
SQL Injection and Its Detection Using Machine Learning Algorithms and BERT
@INPROCEEDINGS{10.1007/978-3-031-28975-0_1, author={Srishti Lodha and Atharva Gundawar}, title={SQL Injection and Its Detection Using Machine Learning Algorithms and BERT}, proceedings={Cognitive Computing and Cyber Physical Systems. Third EAI International Conference, IC4S 2022, Virtual Event, November 26-27, 2022, Proceedings}, proceedings_a={IC4S}, year={2023}, month={3}, keywords={Cyber-attack Security SQL Injection Auto-detection Machine Learning}, doi={10.1007/978-3-031-28975-0_1} }- Srishti Lodha
Atharva Gundawar
Year: 2023
SQL Injection and Its Detection Using Machine Learning Algorithms and BERT
IC4S
Springer
DOI: 10.1007/978-3-031-28975-0_1
Abstract
SQL Injection attacks target the database of applications to extract private information or inject malicious code. In this paper, we attempt to present a well-researched and practiced methodology to detect SQL Injection attacks accurately. These kinds of attacks are a very common means of network security attacks which can cause inestimable loss to the database. Building measures against them is a current research hotspot. Considering the possible complexity of queries involved and the need for a quick and efficient detection system in place, turning to machine learning techniques to combat and detect such attacks is the right choice. This is why we have undertaken the task of analyzing a number of machine and deep learning algorithms on a vast dataset of 41,770 points (consisting of both malicious and normal queries). We aim at finding a system that is accurate and fine-tuned for the best possible results and test each of the algorithms on various performance metrics to identify the one that performs the best. BERT outperforms the rest with a validation accuracy of 99.98%.
