Real-Time Credit Card Fraud Detection and Visualization

This application uses a machine learning model and Spark Structured Streaming to detect credit card fraud in real-time.

Table of Contents

• Prerequisites
• Setup Instructions
• Running the Code
• Visualizing the Data

Prerequisites

Install Docker and Docker Compose

Ensure both are installed on your machine:

• Docker Installation Guide
• Docker Compose Installation Guide

Python Environment

• Install Python 3.11.11
• Install required libraries:

pip install -r requirements.txt

Setup Instructions

1. Start Docker Containers

docker compose up

2. Verify Services Are Running
   Ensure the following services are up:

• cassandra
• zookeeper
• kafka
• fastapi
• dashboard
• react-dashboard

Running the Code

Run the following steps:

1. Initialize the database:

python -m app_initialization.main

(Note: You may need to wait for Cassandra to fully initialize before running this.)

2. Start the Spark Structured Streaming application:

python -m spark-app-v2.structured_streaming_fraud_detection

3. Start the transaction producer:

python -m producers.transaction_producer

4. Set up the frontend dashboard:
   • Navigate to the dashboard-app folder:

cd dashboard-app

• Install dependencies:

npm install

• Start the development server:

npm run dev

Visualizing the Data

Open your browser and navigate to:

http://localhost:5174

to view the real-time fraud detection dashboard.

---