Powered by TensorFlow: How machine learning helps advance access to human rights information
summary
HURIDOCS and UPR Info use BERT on TensorFlow Hub to build a classifier that assigns human-rights topics to UPR recommendations, trained on 60,000 hand-labeled examples and deployed with TensorFlow; the system handles up to 8,000 recommendations a year, uses explicit accept/reject feedback to retrain, cuts database update time from 2–3 months to 1 week, and clears a backlog while improving suggestion accuracy over time.
AI-written summary · about
related talks
15mTensorFlow Hub for real world impact
01mMachine Learning for Human Rights Advocacy
31mOperationalizing Responsible AI
31mWhen humans teach machines
20mHuman-Centered AI for Sustainable Development
41mUsing AI to Build Fair and Equitable Workplaces
11mHuman-centered technology and the role of Responsible AI
21mTransform your web apps with Machine Learning