AI-enabled platform to reduce bias in datasets released
On Wednesday, London-based Synthesized launched a platform to help organizations identify and rectify biases in their data. Synthesized touts the platform as the “first publicly available solution to accurately detect and remove biases in data.” A “freemium” Community Edition of the platform designed to mitigate bias in data is now available.
“The reputational risk of all organisations is under threat due to biased data, and we’ve seen this will no longer be tolerated at any level. It’s a burning priority now and must be dealt with as a matter of urgency, both from a legal and ethical standpoint,” said Nicolai Baldin, CEO and founder of Synthesized in a press release.
Creating more equitable datasets
Overall, the platform allows companies to identify biases in data related to age, gender, race, sexual orientation, religion, and more, according to the company. In a press release, Synthesized listed a series of potential use cases spanning industries including more equitable insurance claim assessment and “fairer” credit ratings in finance. The company said the platform could also be used to reduce the risk of data bias in university admissions and the hiring process.
The platform leverages artificial intelligence (AI) to automatically eliminate data biases once identified, per Synthesized, as part of a process called “rebalancing.” The platform uses an algorithm to incorporate “randomised changes” to the biased data and then create a new “entirely synthetic dataset.” For more “granular control of the rebalancing process” organizations can also change individual attributes in a dataset.
The company said that the system does not require extensive technical expertise or coding experience. To utilize the platform, organizations will need to first upload a data file such as a spreadsheet. Organizations can also connect the platform with databases from services like Google Cloud, AWS, Azure, and more.
After the data has been analyzed, the platform provides a “Synthesized Total Fairness Score” highlighting biased portions of data as well as the percentage of data determined to be biased. In about 10 minutes, the platform is able to “crunch” more than 4 million rows of data, per Synthesized.
“Synthesized’s Community Edition for Bias Mitigation is one of the first offerings specifically created to understand, investigate, and root out bias in data. We designed the platform to be very accessible, easy-to-use and highly scalable, as organisations have data stored across a huge range of databases and data silos,” Baldin said.