New study says encouraging consensus among data labelers from diverse backgrounds can lead to better AI for detecting misogyny and harassment on social media (Khari Johnson/Wired) TECH REVIEW

Khari Johnson / Wired:
New study says encouraging consensus among data labelers from diverse backgrounds can lead to better AI for detecting misogyny and harassment on social media  —  Misogyny on the internet too often slips through the filters of content moderators.  A new method hopes to inject more nuance into the process.

from Techmeme https://ift.tt/2VqjBAd
via IFTTT