Written by Byron V. Ahid
To sell us more products and services, Google, Facebook, and Amazon's algorithms comb through our digital footprint.
There's nothing inherently wrong with companies seeking to better understand their customers. However, over the past two decades, the practice of analyzing user data has not progressed much beyond serving the business models of these tech giants.
I had the opportunity to visit with NTT Research's Senior Researcher in the Cryptography and Information Security (CIS) Laboratory to learn more about the progress being made on a promising concept called “privacy-preserving aggregate statistics.”
Rising data privacy regulations emphasize the need for such capabilities, Boyle told me. And in the long run, the ability to analyze our online behavior in a more inspired and respectful way could yield far greater benefits than just encouraging consumer impulse purchases. There is a gender. Watch the accompanying videocast for a complete drill-down. Here are some important points:
Increased regulation
Tech giants aren't the only ones with a strategic imperative to better understand user behavior. Companies across all industries have long sought to better understand how consumers use their products and services. This guides product improvements, determines future investments, and often shapes the next big innovation.
Our smartphones, wearables, vehicles, and buildings are loaded with sensors that collect detailed information about our daily activities and provide a source of information about what we like and how we behave. Embedded. However, the intensive capture of personal data points in the absence of reasonable oversight is causing consumer anxiety, and rightly so.
This has led to stronger data privacy regulations. For example, Europe's General Data Protection Regulation (GDPR) and California's Consumer Privacy Act (CCPA) are two important laws aimed at protecting consumer privacy in the digital age. Both regulations have significant implications for businesses seeking to collect consumer data and apply aggregate statistical analysis.
GDPR requires companies to establish a legal basis for data processing and to ensure that individual identities are protected through aggregation and anonymization methods. CCPA, on the other hand, focuses on ensuring that personal information is not sold without a consumer's knowledge or against their will.
Partitioning user data
Now, the problem is: Businesses are eager to extract useful insights from user data, but many are losing sight of the fact that it will be much more costly to obtain detailed tracking details going forward. This led NTT Research to explore ways companies could perform aggregate data analysis of consumer data that incorporates privacy, Boyle said.
Privacy-preserving aggregate statistics revolve around dividing sensitive user data into pieces. Although each piece by itself tells us nothing about the original data, meaningful calculations can be performed on the pieces and eventually they can be recombined. Mr. Boyle explained how he could configure his system to privately divide sensitive user data into two segments, thus making his telemetry private.
In addition to meeting compliance, this approach can significantly reduce the costs associated with collecting and storing sensitive personal data and potentially improve revenue, she says. In addition to being ready to develop and deliver technology, Boyle said:
“The goal is to develop a solution that can take personal information and break it up into pieces, so in a sense it never touches the personal data and only learns the collective information,” she says. “The difficult part is designing this partitioning procedure so that you can actually calculate these parts separately.”
greater good
In a world increasingly concerned about data privacy, this new twist on data analysis could help companies comply with privacy regulations and ease consumer fears. It could also provide a way for companies to gain data-driven insights in a more efficient and elegant way.
Boyle believes that companies in all industries, including healthcare, financial services, energy, and consumer goods, can quickly leverage this new approach to extract more useful insights from a growing data lake of somewhat random consumer data. We've pointed out how you can get started.
They will be able to examine the steadily increasing influx of consumer data at a summary level and discover overall patterns and trends. For example, NTT Research successfully tested advanced privacy-preserving calculations using common benchmarking tools such as histograms, mean and standard deviation, maximum and minimum values, and top common values.
It's just a starting point. As this type of advanced encryption moves into mainstream use, it could encourage innovators to leverage their digital footprint for purposes beyond simply tailoring advertising.
For example, in one project, social scientists in Boston applied privacy-preserving calculations to employee pay and benefits data at multiple companies to determine whether there was a pay gap between men and women.
It's not hard to imagine how privacy-preserving statistical analysis could help climate scientists better understand energy usage patterns or medical researchers track the spread of disease.
“If we can somehow combine this information and learn something from around the world, it's very powerful,” Boyle says. “It's very exciting to be in a position where mathematical concepts like abstract algebra actually play a role in the design of logical systems that help solve big problems.”
Change will continue. I will continue to pay attention and report on this.
Ahid
Byron V. Ahid, a Pulitzer Prize-winning business journalist, is dedicated to raising public awareness about how to make the Internet as private and secure as it should be.
(LW will provide consulting services to the target vendor.)
January 29, 2024