Communications of the ACM
This App Offers Personal Recommendations While Keeping Your Data Private
PrIA, which stands for Private Intelligent Assistant, surfaces articles it recommends in one tab and those recommended by Google News in another, and asks users which ones they find useful.
Credit: PrIA
A personalized news recommendation application collects data about the user to suggest relevant stories without sending any personal information to a remote server.
The Private Intelligent Assistant (PrIA), created by researchers at Stony Brook University and IBM, requires both a laptop and an Android smartphone to work. The phone and laptop use data from the user's social media and Google Chrome activity to create a profile that is stored locally on the device.
The laptop app downloads stories from Google News and makes recommendations for which news stories it thinks the user would like to read based on their profile. The app then sends those stories to the user's smartphone.
Researchers had six people install the app on their devices and compared its suggestions to the ones made by Google News when participants logged into their Google account. The app's recommendations were slightly less accurate, with Google News performing about 14% better.
However, the researchers say results can improve by incorporating a wider range of data about the user.
From Technology Review
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found