CLARA: Confidence of labels and raters

This Research paper walks through a full [[ bayesian ]] model for inferring when decisions made by unknown annotators can be trusted.

They use Gibbs Sampling to do inference.