🚔 KSP Hackathon - Crime Intelligence Platform

An AI-powered Crime Intelligence Platform built for Karnataka State Police Hackathon to visualize crime patterns, detect hotspots, analyze criminal networks, and predict future crime risk using GIS and Machine Learning.

---

📸 Project Demo

Dashboard

Crime Heatmap

Criminal Network Analysis

Risk Prediction Dashboard

Ground Truth Recovery

---

✨ Features

🟢 Phase 1 – Crime Mapping & Visualization

• Interactive Karnataka crime map
• Live crime incident visualization
• PostgreSQL + PostGIS integration
• Synthetic crime data generation
• District-wise filtering
• REST APIs using FastAPI

---

🟡 Phase 2 – Hotspot Detection

• Crime Heatmaps
• Station-wise density visualization
• Red Zone Detection using Z-Score
• Day/Night crime filtering
• Weekday vs Weekend analysis

---

🔵 Phase 3 – Criminal Network Analysis

• Modus Operandi (MO) clustering
• Interactive criminal relationship graph
• Force-directed visualization
• Crime pattern discovery
• Cluster-based suspect analysis

---

🟣 Phase 4 – AI Risk Prediction

• XGBoost-based crime forecasting
• SHAP Explainability
• Station-wise risk scores
• Feature importance visualization
• Ground Truth Recovery Dashboard

---

🔴 Phase 5 – Officer Dashboard

• Station Officer View
• SCRB Analyst View
• Role-based dashboard UI
• Improved user experience
• Shared dashboard components

---

🏗️ System Architecture

                    React Frontend
                           │
          ┌────────────────┴────────────────┐
          │                                 │
      FastAPI Backend                 ML Services
          │                                 │
     PostgreSQL + PostGIS         XGBoost + SHAP
          │
   Synthetic Crime Generator

---

🛠️ Tech Stack

Frontend

• React
• Vite
• Leaflet
• React Force Graph
• Recharts

Backend

• Python
• FastAPI

Database

• PostgreSQL
• PostGIS

Machine Learning

• Scikit-learn
• XGBoost
• SHAP
• Pandas
• NumPy

---

📂 Project Structure

KSP-Crime-Intelligence/
│
├── backend/
├── frontend/
├── ml/
├── docs/
├── images/
├── requirements.txt
└── README.md

---

🚀 Installation

Clone Repository

git clone <repository-url>
cd KSP-Crime-Intelligence

Database Setup

createdb ksp_crime

psql -U postgres -d ksp_crime -f backend/schema.sql

Backend

cd backend

python -m venv venv

source venv/bin/activate
# Windows
venv\Scripts\activate

pip install -r requirements.txt

python generate_data.py

uvicorn main:app --reload

Frontend

cd frontend

npm install

npm run dev

Visit:

http://localhost:5173

---

📊 Future Enhancements

• Real-time CCTNS Integration
• Authentication & Authorization
• Live Crime Streaming
• AI-based Anomaly Detection
• Predictive Patrol Route Planning
• Cloud Deployment

---

👨‍💻 Author

Keerthana K

AI & Machine Learning Enthusiast | Full Stack Developer | ML Engineer Aspirant

---

⭐ If you found this project useful, consider giving it a star!