DURHAM, N.C., June 21, 2023 (GLOBE NEWSWIRE) -- Tumult Labs, a differential privacy pioneer, today announced a case study on how the Wikimedia Foundation leveraged the company’s Tumult Analytics platform to publish detailed data about Wikipedia usage patterns without compromising the privacy of users and their location. For the first time, anyone interested in better understanding Wikipedia page view data at a more granular, country-specific level can find regularly published data and useful insights. By leveraging Tumult Labs’ differential privacy solution, the Wikimedia Foundation can share and analyze sensitive data without risking user privacy. To learn more about how this is made possible, read the joint case study.
Differential Privacy Does What Traditional Data Sharing Methods Cannot
An emerging data privacy paradigm, differential privacy is a standard for computations on data that limits the personal information that could be revealed. Traditional methods, such as simply aggregating data, are not enough to prevent re-identification risk, which can expose the identity of users and their data patterns. Instead, differential privacy ensures sensitive data can be shared but with rock-solid privacy guarantees. This is especially vital as sophisticated data privacy attacks are becoming more common and data privacy regulations evolve globally.
The team at Tumult Labs has been at the forefront of differential privacy research and applications since 2008. The company’s flagship product, the Tumult Analytics Platform, ensures customers like the Wikimedia Foundation can gain greater value from sensitive shared data and ensure better privacy risk management.
Tumult Enables a Careful Balance of User Privacy and Open Data Access
The Wikimedia Foundation, the non-profit that operates Wikipedia and other Wikimedia projects, strives to achieve a balance of both open data access and user privacy. Wikipedia, the world’s largest online encyclopedia, has a vast content infrastructure – more than 1.7 billion unique devices access Wikipedia pages and about 300,000 editors contribute to Wikipedia every month. The organization wanted to provide a more granular look at which pages, by topic and by country, garner the most traffic and usage. This data is valuable not only to editors and researchers but to global visitors of Wikipedia’s many pages. As such, the goal was to ensure such data could be fully accessible to the public and yet still align with the Foundation’s strong transparency and privacy principles and policies. For example, location information can be extremely sensitive, and editors or users of delicate topics might not want their location to be revealed, even at the country level. In addition, people browsing Wikipedia rightly expect their behavior on the website to stay private.
The Wikimedia Foundation conducted an in-depth comparison of available open-source differential privacy tools and selected Tumult Labs and the Tumult Analytics platform. A collaborative effort from both organizations, the teams worked to outline the privacy project objectives, quickly test the approach, and then fine-tuned the deployment to ensure success. As a result, Wikipedia editors and others now have access to open data that would have otherwise not been possible to share. The published data provides useful insights about Wikipedia usage patterns while still providing strong privacy guarantees to people browsing the website.
“Data privacy is a complex issue that is fraught with risk. Our team has spent the last 15 years working to understand how to navigate these intricacies and has structured our platform to ensure our customers have the best of both worlds – the ability to share and collaborate with data while ensuring privacy is protected,” said Gerome Miklau, Founder and CEO, Tumult Labs. “We are thrilled to partner with the Wikimedia Foundation and look forward to helping the team further expand their use of differential privacy in the months ahead.”
For more information, check out our blog post detailing the project.
About Tumult Labs
Tumult Labs is on a mission to redefine data privacy protection. Founded by a leadership team that has been at the forefront of differential privacy research and applications since 2008, our company helps organizations share and analyze sensitive data without compromising privacy. Customers, including the U.S. Census Bureau, U.S. Internal Revenue Service, and the Wikimedia Foundation, use the Tumult Analytics Platform to gain greater value from sensitive shared data and ensure better privacy risk management. For more on how we make it happen, visit https://www.tmlt.io/.
