non-binary text
er1p opened this issue · comments
you might consider being more generous with the pronoun search to catch the following variations:
- "pro-nouns" as a synonym for "pronoun"
- the string "they/them" even without the word pronoun, plus "them/they"
OTOH, if you are too generous with the string "pronoun", you might pick up:
- "pronounced 'ree-del' " and similar guidance about name pronunciation
not sure what you would do with:
- "she or they for pronouns"; "she/her/they"; "they/she"; "he/they"
that's all the obvious stuff from my followers, followings
nice app, nice start
Thanks for the feedback! Sorry I didn't notice until now. The code doesn't look for the string "pronoun" in the Twitter profile at all, it only looks for "he", "she", "they" as words:
https://github.com/ajdavis/twitter-gender-distribution/blob/master/analyze.py#L44
So far this seems like the best heuristic I can find for detecting when people declare their pronouns in Twitter profiles.