Reproduce “Fairness in Rating Prediction by Awareness of Verbal and Gesture Quality of Public Speeches” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ankani Chattoraj, Rupam Acharyya, Shouman Das, Md. Iftekhar Tanveer, Ehsan Hoque
The authors measure transcript and facial-gesture variety in TED talks as heterogeneity (HEM), then test its relationship to ratings and group differences. They add a penalty to a rating predictor, encouraging similar predictions for talks with similar HEM. On this dataset, fairness measures improve while accuracy across six labels falls from 67% to 64%; this does not establish fairness generally.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.