Thanks for writing this comment. I can't think of a better way to describe what this library does than reading these super simple examples.

Check out the docs. It is chock full of them.

https://github.com/nlp-compromise/nlp_compromise/blob/master...

Does the same thing work if the name is lowercase?

Stanford nlp won't recognise "I want to go to sydney" with a location because it expects proper nouns to be capitalised for places (and other similar named entities I believe). I tried to make their case less models work but a lack of time and patience resulted in a more specific fix for what we needed.

  > nlp.text('i want to go to sydney.').places()[0].city
  'Sydney'

  > nlp.text('tony hawk did a kickflip.').people()[0].text
  'tony'

So in the first case, it clearly does recognise that Sydney is a city independent of capitalisation. In the second case, it does recognise the first name but not the last, instead tagging "hawk" as a noun. The same happens with "john smith", however "john thomas" is recognised as a full name.

To be fair, I think that my grandmother, not familiar with gaming/skating celebs, would tag it the same way.

Thanks. Why don't more people write docs like this, leading with examples?

It's called README file. It would blink if it could.

I added this to an example chatbot lesson I made for a high school intro CS class.

http://codepen.io/lilgreenland/pen/zqXLeJ?editors=0010

can anyone think of some other cool tricks I can do with this to show my class on Monday?

Look up example questions for the Loebner Prize, and see if you can tweak your bot to answer some of them.