Anonymization of Medical Data Streams using Prink

Overview

This repository contains a demo application for a Master's thesis project focused on the anonymization of medical data streams using Prink. The application consists of dockerized microservices to stream medical data and perform in-transit anonymization.

Architecture Diagram

Key Features

• Dockerized microservices
• Streaming medical data
• In-transit anonymization using Prink

---

Getting Started

Prerequisites

• Docker
• Docker Compose

Build & Run

docker compose up --build

Service	URL	Default Credentials
Prink Overview	http://localhost:8081/#/overview	-
Grafana	http://localhost:3000/login	admin / admin
Prometheus	http://localhost:9090/query	-

Note: To log in to Grafana, use credentials admin/admin, then choose skip.

...When finished

docker compose down

Dashboard Views

Configuration

Privacy Parameters

Parameter	Default Value
k	5
l	3
delta	125
beta	50000
zeta	10000
mu	100

Note: The privacy parameters can be configured in the GangesEvaluation.java Prink Job file.

Prink Kafka consumer properties

• bootstrap.servers: Kafka broker addresses.
• group.id: Consumer group ID.
• key.deserializer: Deserializer for the message key.
• value.deserializer: Deserializer for the message value.

Note: The consumer properties (connecting Prink to other topics) can be configured in the GangesEvaluation.java Prink Job file.

Flink Configuration

• The Flink configuration file is located at flink-conf.yaml.
• Key configurations include:
  • jobmanager.rpc.address: Address of the job manager.
  • taskmanager.numberOfTaskSlots: Number of task slots per task manager.

Third-Party Licenses

This project uses third-party components:

• Apache Kafka, Zookeeper, Flink, Prometheus (Apache 2.0)
• Grafana (AGPL v3)

These components are used as-is via container images.

License

This repository’s own source code is licensed under the MIT License.