Since the official Twitter Website is not the only possiblity to send tweets, the original DataTable (ALL the europe-wide data) was queried to determine the major sources & platforms like for example foursquare. This helps with finding out which platforms are the most often used to send twitter texts.
The queries are written in a very
simple style - the following one, just as an example, gathers all the
"foursquare - sent" tweets:
select source from geo_tweets_europe_feb2012 where source like '%foursquare%';
Since
we were only interested in the number of tweets sent per source, this method worked fine for us.
The "tweet-sources", are basically stored in a speperate column in the datatable and contain the
information on the source of twitter texts. “foursquare” is the
source against which we query the database in this example case. Using
this method, the following results were obtained for the different
(major) sources of tweets in our data:
Twitter: 11,221,660 rows
foursquare: 1,855,073 rows
Tweetdeck: 86,432 rows
Tweetcaster: 121,134 rows
dlvr.it: 89,369 rows
facebook: 310 rows
instagram: 151,478 rows
Since we were interested mostly in the tweets concerning financial talk, the procedure was repeated for the according data-selction:
Twitter: 227,701 rows
foursquare: 42,994 rows
Tweetdeck: 2,404 rows
Tweetcaster: 2,539 rows
dlvr.it: 7,471 rows
facebook: 5 rows
instagram: 3,257 rows
As
we basically expected, Twitter itself seems to be the main
sources of the tweets, but foursquare's strong bonding is clearly
visible when looking at the distribution above. Other sources like
Tweetdeck, Tweetcaster, the Website dvlr.it or even instagram contribute
with rather small numbers.
No comments:
Post a Comment