Our analysts have created a comparative word cloud that compares and represents the top two twitter hashtag discussions surrounding the current Julia Gillard and Kevin Rudd leadership debacle. The image shows the top terms mentioned in Twitter conversations that include the hashtags #kevenge and #respill, and then overlaps them to look for key themes.
By quickly scanning this comparative word cloud, we can see what terms or words are being used in volume in relation to the specific hashtag. At the time of creating this at 5:45pm on 23rd of Feb, it’s clear that Kevin Rudd’s shock resignation and media appearances were overshadowing and dominating conversation when compared to Julia Gillard.
Users discussing #kevenge were also mentioning “Australian Labor Party”, as well as policy issues which Kevin Rudd’s known for, such as ‘economy’ and ‘tax’.” Users also mention ‘leadership’, a topic that has gained some significance as the K-Rudd has stressed that he has the leadership capacity to win the next election.
Interesting to note that when users discussed “#respill” (the leadership spill yet to occur on February 27th), users were mentioning ‘Auspol’ (short for ‘Australian Politics’) as well as mentioning ‘Latikambourke’ (a political and social media reporter at Parliament house) in large volume. Tweeters were also quite active in mentioning ‘Rudd’ and ‘Abbot’, but in comparison tweeted about ‘Gillard’ in low volume.
Most of you will hopefully be familiar with word clouds. Word Clouds are visual representations of the number of times each word in a transcript is used. The higher the frequency, the larger the word is printed. Word Clouds are often used to quickly visualize the content of a specific text.
There are word cloud generators everywhere online, such as Wordle where you enter in all of your words, and it will show you the cloud almost instant of the quickest ones.
Then there’s Twitter Word Cloud Generators such as ouTwitMe, where you enter in a Twitter search term or Hashtag, and it will show you the most mentioned words that have been tweeted that include that term/hashtag.
Currently as I write this, Kevin Rudd and Julie Gillard are all over the news (and Twitter) with regards to the Australian Labour Party leadership, and the top trending hashtag in Australia is “#respill”. Here’s a Word cloud of the top 50 Tweets that include the term “#respill”.
The 2nd similar Twitter hashtag currently trending “#Kevenge”, a word cloud of the top 50 words of that hashtag can be seen here:
The comparative Word clouds look to combine the above two charts and look for key themes.
Drew Conway came up with the idea of a comparative wordcloud, using the horizontal axis to show how two sets of text compare. Like other wordclouds, the frequency of a particular word is represented by the size of the word in the chart. A comparative word cloud goes further by showing how a particular word is weighted between the two things being compared.
The coding was all done by Drew Conway who made it freely available under the Simplified BSD Licence. We copied the code into “R” and simply set the parameters we wanted. (“twitter_word_cloud.R” – full credit to Drew Conway - Copyright (c) 2011, under the Simplified BSD License. All rights reserved.)
Besides the possibilities of tracking the ever so exciting political movement of our nation, comparative word clouds can offer us the ability to quickly track and create accurate trends for our clients and brands we manage, using data that’s coming straight from the user.
For example: we’re able to use comparative word clouds to track social mentions of brands we manage against our competitors, we can keep our finger on the pulse of any social media crises that arise and deploy strategy, we can even figure out who the most popular Visual Jazz Isobar tweeter is in the office!
As nerdy as this sounds, this is really exciting stuff! The possibilities to gain insights and create user focused trends are endless.