Advanced Malware Detection Platform
Overview The Metamal-D-Alert system is a comprehensive, AI-driven malware detection platform designed to analyze files using a combination of dynamic and static analysis techniques. It leverages modern deep learning models to provide accurate and efficient malware detection results. This project integrates a user-friendly frontend built with React.js, a robust backend powered by Node.js/Express.js, and a deep learning model deployed within Docker containers for consistent performance across environments.
Features User Interface (UI): Developed with React.js to provide a responsive and interactive platform where users can upload files for analysis and view results. Backend Server: Built with Node.js and Express.js to handle file uploads, communicate with the hybrid analysis API, and interact with the deep learning model. Hybrid Analysis API: Combines dynamic and static analysis to generate a detailed JSON report, which the deep learning model processes to detect malware. Deep Learning Model: Utilizes a CNN-BiGRU model, enhanced by NLP embeddings, for accurate malware detection. Deployed using Docker for environment consistency. Database: NoSQL database for storing analysis results and logs, ensuring scalability and flexibility. System Architecture The system is composed of several high-level components:
User Interface (UI): React.js Backend Server: Node.js, Express.js Hybrid Analysis API: Python-based custom API for dynamic and static analysis Deep Learning Model: CNN-BiGRU with NLP embeddings, deployed in Docker Database: MongoDB for storing results and logs
Implementation Details Frontend: Implements a smooth drag-and-drop file upload experience using React Dropzone. Displays analysis results dynamically with React state management. Backend: API integration handled through Node.js middleware, ensuring secure and efficient communication between the frontend, hybrid analysis API, and deep learning model. Hybrid Analysis API: Executes files in a sandbox environment for dynamic analysis while performing static code inspection to provide a comprehensive threat assessment. Deep Learning Model: The model is trained on a labeled dataset and fine-tuned with hyperparameter adjustments. It combines CNN for feature extraction and BiGRU for sequence analysis. Performance Metrics Accuracy: 98% on the test set. Response Time: Average of 5 seconds for file analysis. Scalability: Efficiently handles concurrent file uploads and analyses. Testing Unit Testing: Ensured individual components, such as file upload and API endpoints, function correctly. Integration Testing: Verified seamless interaction between the frontend, backend, hybrid analysis API, and deep learning model. Acceptance Testing: Conducted with end-users to validate overall system performance and usability. Getting Started Prerequisites Node.js Docker MongoDB Installation Clone the repository:
git clone https://github.com/yourusername/metamal-d-alert.git cd metamal-d-alert Install dependencies:
cd test npm install
cd ../server npm install Start the backend server:
cd server node index.js Start the frontend server:
cd test npm start Run the Docker container:
docker
docker build -t .
Usage Access the application via http://localhost:3000. Upload a file for malware analysis. View the detailed analysis report generated by the hybrid analysis API and deep learning model. Contributing Contributions are welcome! Please fork this repository and submit a pull request with your updates.