Binary Classification Research Project with GNN 📊🧠

I have recorded some of my work here in this vedio with our Prof Dr. Muhammad Masoom Alam(mailto:[email protected])

Video Overview 🎥

In this vedio, you'll find informative content related to binary classification and AI-driven security research. Click the image above to start watching!

Project Overview 🌐

Welcome to the Binary Classification Research Project repository! This repository represents a journey of exploration and innovation in the realm of binary classification using Graph Neural Networks (GNN). Our goal is to enhance accuracy and interpretability in binary classification, drawing inspiration from the DeepReflect: Discovering Malicious Functionality through Binary Reconstruction. paper and utilizing the power of GNNs and the XAI. But DR used auto-encoder approach and we are using GNN instead , they have used binary ninja disassmebler and we used idapro, r2pipe etc, PE python libraries etc .

Contents 📂

• cfg_generator_ida.py: Script to generate control flow graphs (CFG) using IDA Pro for input binary data.
• testing.py: Comprehensive script with various implementations, including r2pipe methods to convert assembly to CFG, along with other tools for transforming assembly or executable files to CFG.
• Linux_json.json and windows_json.json: JSON files containing sample data (private) for Linux and Windows environments.
• README.md: You're currently reading this file!

Binary Classification Research Project Resources 📚🔍

This repository serves as a curated collection of research papers, tools, and references related to binary classification, malware detection, and AI-driven security.

Research Papers 📄

• Static Disassembly of Obfuscated Binaries
• Observational Approaches to Malware Detection
• Deep Android Malware Detection and Classification
• PE-Miner: Mining Structural Information to Detect Malicious Executables in Realtime
• AVclass: A Tool for Massive Malware Labeling
• Data Mining Methods for Detection of New Malicious Executables
• Semi-Byte N-Gramas for High-Speed Malware Detection
• Obfuscation Benchmarks
• FunctionSimSearch: Tool for Code Similarity Comparison
• HDBSCAN: Hierarchical Density-Based Clustering
• Kaggle Malware Classification Challenge
• Malware Detection Using Machine Learning
• Autograph: Automated Distributed Worm Signature Generation

Tools and Repositories 🛠️

• Unipacker: Unpacker for Packed Executables
• Nucleus: An End-to-End Classifier for Binary Executables
• Awesome Executable Packing: Curated List of Packing Tools
• FunctionSimSearch: Tool for Code Similarity Comparison

Conferences and Journals 📰

• IEEE Symposium on Security and Privacy
• USENIX Security Symposium

Disclaimer & Call to Collaborate ❗🤝

The materials you find in this repository offer a glimpse into the ongoing research journey. These snippets are just a fraction of the work we're doing. Much of the project remains private as we continue to refine our approach.

Join the Journey: If you share our passion for leveraging GNNs for binary classification, we invite you to collaborate with us. Together, we can delve deeper into the potential of GNNs and contribute to pushing the boundaries of binary classification.

How to Contribute 🌟

1. Reach out to me at [email protected] to express your interest and intention to collaborate.
2. Upon joining, you'll gain access to the full range of notes, research materials, and our complete approach.
3. Collaborate, innovate, and contribute to the evolution of binary classification with GNNs.

Contact Us 📬

Feel free to contact me via email at [email protected] for any inquiries, suggestions, or if you're excited to embark on this research journey with us.

Thank you for your interest in our ongoing research project. Let's explore the potential of GNNs together! 🚀🧠