Based on the search keyword "tude" and what follows it: (, ) or ?, I
made the same guess myself, but I did not want to assume anything.
Interesting link. I used to do some of that to analyze posts on a
forum in the early days of the internet (to weed out impostors), but
it did not work well.
On Jul 23, 12:42 am, Bjoern <bjoer...@googlemail.com> wrote:
> On Jul 22, 8:49 pm, Joseph <northwest...@gmail.com> wrote:
>
> > That's what I meant. Short of doing a search, with tude[]=%3A) and
> > store it in my cache (which will eat up a lot of API calls), do you
> > have any hints on how to extract this out of the API?
>
> Isn't it just searching for the Strings ":-(", ":-)" and "?"? I don't
> think the attitude detection is more sophisticated than that at the
> moment?
>
> Otherwise, maybe have a look at this:http://danzarrella.com/tweetpsych.html
>
> Although don't put too much weight on it, I guess. But it's a fun
> approach.
>
> Björn
