Introduction¶
Rory is an elastic, cloud-native platform designed to facilitate secure, distributed data mining on encrypted data. It ensures data privacy using standard, differentially private, and post-quantum cryptographic (PQC) methods without requiring data decryption during the analysis process.
⭐ Key Features¶
- Privacy-Preserving Data Mining (PPDM): Execute algorithms like Secure K-Means and Secure KNN directly on encrypted datasets.
- Post-Quantum Security: Integrated support for PQC to future-proof data privacy against quantum computing threats.
- Elastic Architecture: Dynamic scale-in and scale-out of worker nodes based on real-time processing demand.
- Distributed Processing: Automated data segmentation, load balancing, and secure transport pipelines via a Cloud Storage System (CSS).
⚙️ Architecture¶
The platform maps traditional Data Owner (DO) and Service Provider (SP) entities into three core components:
-
Client: Prepares, encrypts, and segments data for externalization.
-
Manager: Orchestrates PPDM tasks, manages storage buckets, and handles load balancing.
-
Worker: The execution engine that applies PPDM algorithms to encrypted chunks.
For a detailed breakdown of the system components and the elastic interaction model, please refer to the Architecture Documentation.