Data are being generated at an unprecedented rate. Over the past year, however, major international corporations like Equifax, Uber, Yahoo and Facebook have been associated with a data breach or leak that has enabled third parties to illegally access millions of personal records. This leaves individuals highly vulnerable to identity theft, financial hacking and other invasions of privacy.
The case that has recently captured headlines pertains to Cambridge Analytica’s data mining activity on Facebook as early as 2014. Facebook, which has unlimited access to millions of user profiles, allowed Cambridge Analytica to conduct targeted advertising campaigns based on personal data.
In some ways, this was predictable. The “Terms of Services” for all of the consumer applications that millions of people use on a daily basis make it clear that the owners (data providers) of these applications can:
- Share data with third parties;
- Use the data to conduct in depth data analytics on demographic populations, which becomes fodder for marketing/advertising agencies; and
- Store the data in any jurisdiction they choose.
Since many data providers can share this data, they are doing so via USB keys, email and file transfer protocol (FTP) sites where large data sets are uploaded. This is precisely what malicious third parties are looking for: low security access points for private data.
York University to the rescue
A team of York researchers, led by Professor Marin Litoiu of the Lassonde School of Engineering and Faculty of Liberal Arts and Professional Studies at York University, wanted to put a stop to this. With the support and guidance of York’s innovation office (Innovation York) and their commercialization partner, MaRS Innovation, the team created a privacy-protected, data-sharing platform that allows data providers to share access to data in a secure manner without releasing raw data or making copies of them.
This platform being commercialized by Bitnobi, Inc. enables the data provider to control access to virtualized segments of data so that end users (business analysts, data scientists or researchers) can choose the data they want to work with in building a data query, instead of acquiring a data provider’s entire data set.
How is this possible? The end user leverages a simple, easy-to-use interface so that she/he can build and launch data queries quickly, based solely on a preview of the data schema with select data records, on the data provider’s infrastructure. The end result of a Bitnobi data job created by the end user is transformed, aggregate data that helps in further analysis or prediction activities.
Federal support leads to future testing
This work has been supported at the national level, which is critical for early-stage innovations emanating from Canadian universities.
Specifically, the technology secured funding from sources such as:
- The Natural Sciences and Engineering Research Council of Canada’s Idea to Innovation Phase 1 and Phase 1b program allowed the York team to build an interface and create the first prototype of the technology, which would later be transferred to Bitnobi.
- The Federal Government of Canada prequalified this technology through their Build in Canada Innovation Program.
With the Build in Canada Innovation Program, the Bitnobi team expects to begin testing their privacy-protected solution in 2018, with the primary goal to demonstrate how source data can remain on premise and facilitate secure access to end users.