RANCANGAN SISTEM DETEKSI DAN PENCEGAHAN SERANGAN SQL INJECTION DAN CAS PADA WEBSITE LOCALHOST

Abstrak

This study designed a SQL Injection and Credential Access Stealer (CAS) attack detection and prevention system on localhost-based websites. SQL Injection and CAS are two forms of malicious attacks that can lead to sensitive data leaks and system compromise. The methods used include a combination of machine learning-based detection, automatic detection frameworks, and the implementation of proxy logic and knowledge base. Experiments were conducted using public datasets and local simulation sites. The results showed that a hybrid approach with the integration of supervised learning and synthetic data was able to improve detection accuracy by more than 99%.

Penelitian ini merancang sistem deteksi dan pencegahan serangan SQL Injection dan Credential Access Stealer (CAS) pada website berbasis localhost. SQL Injection dan CAS merupakan dua bentuk serangan berbahaya yang dapat menyebabkan kebocoran data sensitif dan kompromi sistem. Metode yang digunakan mencakup kombinasi deteksi berbasis pembelajaran mesin, framework deteksi otomatis, serta implementasi logika proxy dan knowledge base. Eksperimen dilakukan menggunakan dataset publik dan situs simulasi lokal. Hasil menunjukkan bahwa pendekatan hibrida dengan integrasi supervised learning dan data sintetik mampu meningkatkan akurasi deteksi hingga lebih dari 99%.