SQL Injection Detection using Machine Learning

A comprehensive machine learning-based solution to detect and prevent SQL injection attacks. This project utilizes multiple models, including Random Forest, Gradient Boosting, and Neural Networks, reaching up to 97% accuracy in identifying malicious SQL queries.

Overview

SQL injection remains one of the most common web vulnerabilities. This project provides:

• High-Accuracy Detection: Trained on over 100,000 queries.
• Multiple Models: Compare performance between Random Forest, Gradient Boosting, and CNN/LSTM.
• Web Interface: A built-in Django web application for real-time testing.

Tech Stack

• Language: Python 3.x
• Web Framework: Django
• Machine Learning: Scikit-learn, TensorFlow/Keras
• Data Processing: Pandas, NumPy
• NLP: TfidfVectorizer

Performance

Model	Accuracy
Random Forest	~97.5%
Gradient Boosting	~97.2%
Neural Network	~96-99%

Project Structure

• webapp/: Django web application for real-time query detection.
• Dataset/: Training and testing datasets (sqli-extended.csv).
• sql_injection_attack.ipynb: Jupyter notebook for model training, EDA, and evaluation.
• randomforest.pkl: Pre-trained Random Forest model.
• gradientboost.pkl: Pre-trained Gradient Boosting model.
• SQl_injection_detection.h5: Pre-trained Neural Network model.
• Testing_queries.txt: Sample queries for testing the models.

Installation

1. Clone the repository:

   git clone https://github.com/your-username/SQL_injection_detection.git
   cd SQL_injection_detection

2. Install dependencies:

   pip install django pandas scikit-learn tensorflow wordcloud seaborn matplotlib

3. Run the Web Application:

   cd webapp
   python manage.py runserver

   Visit http://127.0.0.1:8000/ to test queries.

Usage

1. Open the web interface.
2. Enter a SQL query in the input field.
3. Click "Submit" to see if the query is classified as a "SQL Injection Attack" or "Not SQL Injection Attack".

Dataset Source

The models were trained using a dataset from Kaggle: SQL Injection Dataset.