Calicon09 Word Cloud
July 1 2009, 3:13pm
I created a word cloud for the #calicon09 tweets with Wordle.
Thanks to Yankee in Canada for the idea. Here’s how I did it: 1. Used the twitter API to access the tweets with the #calicon09 hash tag in a json format. 2. Based on some code found around the interwebs, I put together some php for processing the json format. <?php $json = file_get_contents(”http://search.twitter.com/search.json?q=#calicon09&rpp=100&page=1″, true); //getting the file content $decode = json_decode($json, false); //getting the file content as array foreach ( $decode->results as $results ) { echo “{$results->text}\n”; } echo “<pre>”; print_r($decode); $count = count($decode); print_r($count); echo “</pre>”; $text = $array['results']['text']; print_r($text) ?> <?php $json = file_get_contents(”http://search.twitter.com/search.json?q=#calicon09&rpp=100&page=1″, true); //getting the file content $decode = json_decode($json, false); //getting the file content as array foreach ( $decode->results as $results ) { echo “{$results->text}\n”; } ?> Note the api for twitter only returns certain numbers of results per page. I had to run a few pages at 100 results per page to get all the tweets. I know there are ways to automate this, but it’s a little beyond me at the moment, and this was fine for this particular task. 3. Drop the text of all the tweets into a text editor. 4. Shaped the data. I removed the tag #calicon09, removed @usernames, and a few other stop words. I’m sure I could remove more, but I think the resulting cloud is nice. 5. Paste the text into Wordle’s full text form. I’m sure people may find some problems with the way I cobbled this together. I welcome constructive comments.
