Updated code and files from my work on the Stanford Politeness Corpus
Here's a link to some of that work: https://github.com/sudhof/politeness (original) https://github.com/sudhof/politeness/tree/python3
Many thanks to the original team and to Ben for their work here!
@MeyersBS
This work emerged from my exploration of data on politeness to fit in with dissertation on game-theoretic modeling of requests.
I had trouble getting this to work at first. I hope you profit from my initial failures. I'll update this more as I go, with instructions for dummies like myself!
The original classifier prints two probabilities for every sentence as follows:
{"Could you please play some Sigur Ros?" : [.7, .3] }
I'm not great with messy outputs, I've rewritten some of their classifier's code to make the output a little prettier. Hope you like it!
{"Could you please play some Sigur Ros?" : .7 }