Hit the Search API several times per hour and extrapolate results
given the average time between the 100 statuses returned. Or, use the
track resource in the Streaming API to do the same. There are
technical advantages and disadvantages to each approach, but the
result will be roughly the same.
An approximation is all you'll get from any approach, as there is no
totally unfiltered source of statuses -- for example, protected users'
statuses are unavailable, how do you account for deleted statuses,
etc. etc., so you are always estimating.
Services, Twitter Inc.
On Jul 16, 7:27 pm, mclovin <hanoo...@gmail.com> wrote:
> Hey all,
> Basically I wish to get the total number of tweets that contain a word
> each day. In the JSON or Atom they dont give the total amount of hits
> that the query returns. I was wondering if there was a way to do this
> w/o simply taking in the entire twitter stream and and manually
> counting. Perhaps another twitter search engine or something?