Akritimehta01/Shillbid.ai

[](https://bolt.new/~/sb1-ko7gn12b) # ShillBid AI An ML-powered auction fraud detection platform that identifies suspicious bidding behavior, collusive bidder networks, and shill bidding patterns using anomaly detection, graph analytics, and real-time risk scoring. ## 🚀 Features * 🔍 Fraud & shill bidding detection * 📊 Real-time auction risk scoring * 🧠 Machine learning-based anomaly detection * 🕸️ Bidder-seller network graph visualization * 📈 Interactive analytics dashboard * ⚡ Live fraud alerts using WebSockets * 🧾 Explainable AI insights with SHAP * 🗂️ Synthetic auction data generation ## 🧠 Tech Stack ### Frontend * React / Next.js * Tailwind CSS * Plotly / Recharts / D3.js ### Backend * FastAPI * Python ### Machine Learning * Scikit-learn * XGBoost * Isolation Forest * Pandas & NumPy * SHAP ### Database * PostgreSQL ### Graph Analytics * NetworkX * Neo4j (optional) ## 📌 Problem Statement Online auction platforms are vulnerable to fraudulent bidding practices such as: * Shill bidding * Bid inflation * Seller-bidder collusion * Circular fraud rings * Last-minute fake bid pressure ShillBid AI aims to detect and analyze these behaviors using machine learning, anomaly detection, and graph-based fraud analysis. ## 📊 Core Modules ### 1. Fraud Detection Engine Detects suspicious bidding behavior using: * XGBoost * Random Forest * Logistic Regression * Isolation Forest ### 2. Graph Fraud Analytics Builds bidder-seller interaction networks to detect: * Fraud rings * Dense collusive clusters * Repeated bidder patterns ### 3. Explainable AI Provides interpretable fraud predictions using SHAP values. ### 4. Real-Time Alert System Streams live bid events and triggers fraud alerts dynamically. ## 🖥️ Dashboard Features * Fraud risk monitoring * Auction analytics * Suspicious bidder tracking * Network visualization * Live fraud alerts * Explainability panel Frontend setup: cd frontend npm install npm run dev ## 📂 Project Structure shillbid-ai/ │ ├── backend/ │ ├── app/ │ ├── models/ │ ├── routes/ │ ├── ml/ │ └── data/ │ ├── frontend/ │ ├── components/ │ ├── pages/ │ └── dashboard/ │ ├── datasets/ ├── notebooks/ ├── requirements.txt └── README.md ## 📈 Future Improvements * Graph Neural Networks (GNNs) * Real auction API integrations * Advanced fraud ring detection * Cloud deployment * Multi-model ensemble learning * User authentication & role management ## 📜 License MIT License ## 👩‍💻 Author Akriti Mehta

标签：自动化攻击