Technology

Overview

Xafe’s privacy-enhancement technologies with differential privacy algorithms offers robust protection through both Global and Local Privacy Models. Whether you need centralized control or decentralized privacy at the source, Xafe ensures that your data remains secure and compliant with the highest privacy standards.

Global / Centralized Privacy Model diagram showing untrusted third party users, trusted curator with noise and query, raw private data, and data generators

Global / Centralized Privacy Model

Global Privacy Model is a centralized approach that privatizes and adds noise to data queries to protect individual privacy in large datasets.

Technical Overview

In Global Privacy Model, a trusted central entity holds the entire dataset and processes queries by adding calibrated noise / privacy to the results before sharing them. This ensures that the inclusion or exclusion of any individual’s data has a minimal impact on the overall analysis, providing strong privacy guarantees while maintaining data utility.

Business Case

Global Privacy Model is ideal for organizations that manage large, sensitive datasets, such as healthcare institutions or government agencies conducting censuses. It enables them to share insights and statistics without compromising individual privacy, fostering trust and compliance with data protection regulations.

Local / Decentralized Privacy Model diagram showing data perturbed at the source before being shared with an untrusted aggregator

Local / Decentralized Privacy Model

Local Privacy Model is a decentralized approach where individuals add noise & privacy to their own data before sharing it, ensuring privacy at the source.

Technical Overview

In Local Privacy Model, each individual perturbs their data locally by adding noise / privacy before it is collected by any external entity. This approach provides privacy guarantees even when the aggregator or data collector is not fully trusted, as the raw data is never exposed.

Business Case

Local Privacy Model is particularly useful for businesses collecting data from distributed sources, such as mobile apps or IoT devices. It allows companies to gather useful insights while ensuring that individual users’ data remains private, making it an excellent choice for privacy-conscious consumer applications and industries with stringent data protection requirements.