Sign In

Communications of the ACM

ACM TechNews

Bad News for Fake News: Rice Research Helps Combat Social Media Misinformation


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
'No Fake News allowed' sign.

Rice University researchers have discovered a more efficient way for social media companies to keep misinformation from spreading online, by using probabilistic filters trained with artificial intelligence.

Credit: srcmarion.org

Rice University researchers have invented a more efficient method for preventing the online spread of misinformation on social media, using Bloom filters trained with artificial intelligence.

Rice's Anshumali Shrivastava and Zhenwei Dai utilized machine learning to show their Adaptive Learned Bloom Filter (Ada-BF) needs 50% less memory to equal the performance of learned Bloom filters.

According to Shrivastava, Ada-BF’s reduced need for memory means additional capacity for real-time filtering systems.

Shrivastava said, “We need half of the space, so essentially, we can handle twice as much information with the same resource.”

From Rice University
View Full Article

 

Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


 

No entries found