The Art and Tech of Generating Secure and Authentic Synthetic Data

Explore the intricacies of generating secure and authentic synthetic data in this 28-minute conference talk presented by Andrew Carr from Scott Logic and Paul Groves from Citi. Delve into the concept of Data Helix and understand the differences between redaction, anonymisation, and synthesis techniques. Learn about re-identification risks associated with redaction and the importance of differential privacy. Examine a typical synthetic data flow and discover various generic use cases, along with guidance on which approach to consider for specific scenarios. Gain insights into different data generation approaches and explore specific examples. Get a glimpse of future developments for DataHub/DataHelix and find out how to engage further if the topic piques your interest.

Intro
What is Data Helix
Redaction, Anonymisation and Synthesis
Redaction - Re-identification risk
Differential Privacy
A typical synthetic data flow
The generic Use Cases
What approach to consider for what Use Case
Data Generation approaches
A few Specific Examples
What's next for DataHub/DataHelix
If any of this sounds interesting