How do you aggregate annotators when their performance depends on which subset they're labelling? - Githubissues

MatthewJA / lairs

Notes and issues for my PhD

1 stars 1 forks source link

How do you aggregate annotators when their performance depends on which subset they're labelling? #143

Closed MatthewJA closed 5 years ago

MatthewJA commented 5 years ago

I wrote up a derivation of this: https://github.com/MatthewJA/lairs/commit/155a8ae03016d10bec05507cb5e75d883a9088b9