Publishing Open Data? You Need Perturbation

Privacy is more than just Personally Identifiable Information

I know my neighbour's health records are in an open dataset...

I know that she has three children and is in her mid 30s. By combining these facts, I can narrow down the results and easily locate her in the dataset.

Once I have located her...

It's easy for me to discover confidential information about her medical history.

You Need Perturbation

To protect the people in your data, you need a robust confidentiality solution that automatically applies statistical disclosure control to everything you release.

We have worked in collaboration with government agencies around the world to develop a unique data confidentiality solution–perturbation–that addresses privacy concerns without compromising the flexibility or usefulness of your data.

Adjusts Cell Values

Perturbation automatically adjusts cell values slightly so it’s impossible to identify an individual.

Adds Smart Noise

Smart enough to know exactly where to add the noise so as not to affect the usefulness of the results.


Specific values always get perturbed the same way, regardless of how you build the query/table.

Detailed Queries

Safe and secure access to detailed low level data queries.

The Right Balance

No need to choose between maximum protection or maximum data: perturbation is designed to keep your data as safe as possible and as useful as possible.

Fully Automated

Automatically applied on-the-fly to user queries. No subjective judgements. No need to manually adjust the results.