Ads fraud detection using Kafka and Flink

---

Project goals

The goal of the project is to:

• Create a Flink application which will read from Kafka clicks and displays queues
• Detect some suspicious/fraudulent activities
• Output the suspicious events into a file.

Setting up project

Please get into the docker folder and run the docker-compose file.
To do so, simply execute these lines in the same directory as the docker-compose.yml file:

• docker-compose rm -f
  
• docker-compose up

By running this docker-compose, it will :

• Create a Kafka cluster with two topics : “clicks” and “displays”
• Launch a python script which will send events to those two topics and display them in the terminal

This generator simulates few suspicious/fraudulent patterns that you should detect using Flink. There are 3 distincts patterns we want you to find.