hyandell
commented
6 years ago We did some light research on this internally and it doesn't look like simple analysis works. Maybe deeper learning that understands what a happy and a sad issue is, but that'll need manual annotation to create.
Add a report to show happiness etc in issues/PRs.
This type of thing:
https://github.com/stathack/Presidential-Semantic-Analysis/tree/master/opinion-lexicon-English