Use NLP to detect nationality statistics

[jayvdb]

Use the participants display name to determine nationality.

Discard data which is unsuitable or confidence level is low, but clearly display what percentage of the participants display names were unsuitable or was unusable for the statistics.

wrt unsuitable data, if #32 doesnt land first, skip any display name which contains numerals and doesnt contain an uppercase letter.

To warm you up a bit, see here

Come onto the coala Zulip to discuss how to solve this.

[andrewda]

This is really cool! If #8/#32 lands first, we'll also be able to use a user's location from GitHub.

[invisible-defects]

Sounds really interesting! Have a couple ideas on this, will share them in Zulip soon.

[jayvdb]

This task was revised slightly, and wont be done using NLP.