What is Ant Privacy-Preserving Computation Platform
Ant Privacy-Preserving Computation Platform (AntPPC) is a privacy-preserving computation service that uses technologies such as federated learning (FL), Secure Multi-party Computation (MPC), Private Set Intersection (PSI), trusted execution environment (TEE), and differential privacy (DP). The service analyzes and extracts value from data while protecting private information.
Ant Privacy-Preserving Computation Platform has five components: Federated Modeling, Secure Multi-party Modeling, Secure Multi-party Analytics, Private Set Intersection, and the Management Console. These components provide joint data analytics, private set intersection, model training, model evaluation, model prediction, and centralized resource management. The platform is peer-to-peer, open, and security-first. It desensitizes and decentralizes key information and uses end-to-end encryption for data transmission. This process ensures that data is usable but not visible. The platform meets the stringent requirements for security, compliance, and privacy protection in data collaboration. It creates a trusted and fair environment for all parties to process and use data. This helps promote business growth and innovative applications.
Benefits
Professional
The platform's privacy-preserving computation technologies comply with the 'Technical Specification for Secure Multi-party Computation in Financial Applications' issued by the People's Bank of China. The platform was also one of the first to be certified by the National FinTech Evaluation Center for financial application technology.
Secure
The platform integrates multiple security solutions, such as cryptography, machine learning, and encrypted hardware. This approach breaks down data silos without exposing sensitive data from any party. Data is usable but not visible, meeting security and compliance requirements.
Efficient and self-service
The platform provides a one-stop, end-to-end solution for developing application services. You can use visual interactions to analyze and extract value from data.
Components
User | Component Name | Function |
|---|---|---|
Institutional resource administrator | Centrally manage core resources and configure projects for the Ant Privacy-Preserving Computation Platform. | |
Data application developer | Develop, evaluate, iterate, and update models without moving raw data from its original location. | |
Train and predict models collaboratively by exchanging information. Raw data is not moved from its original location. | ||
Analyze data from multiple parties using secure encryption. Raw data is not moved from its original location. | ||
Find and analyze the intersection of datasets from multiple parties without revealing any additional information. |