I was interested to see how people on twitter were reacting to the Government’s carbon announcement. The government has been taking a bashing in the polls, but I wanted to see how things were looking from a social media perspective.
To perform this analysis I used R an awesome stats language, a ‘sentiment-lexicon’ from Hu & Liu and the method described in this powerpoint by Jeffrey Breen.
Essentially what I did was download the most recent 1,500 tweets that contained the phrase ‘carbon tax’ and cross-referenced them against two lists of ‘positive’ and ‘negative’ words. Then, each tweet was given a score according to how positive or negative it was. The more positive words, the higher the score.
The following graph shows the spread of people’s reactions:

So what does this mean?
As you can see, the highest number of people (575) have had a neutral reaction i.e. their tweet contained neither positive nor negative words.
The reaction on the positive side is stronger than on the negative. More people seem to be passionately positive about the carbon tax than passionately negative. If you just look at those people that have had a strong reaction (very positive or very negative) the reaction is pro carbon tax with 219 very positive, and only 125 very negative. This may be good news for the government.
However in total the ledger is quite even, with slightly more people negative than positive. There were 492 tweets that were negative, and 433 that were positive.
To give you an idea of the sort of things both the very positive and very negative tweeters have been saying, I made these two tag clouds (ooh pretty
). Unsurprisingly, fear and anger feature heavily on the negative side and much kinder emotions on the positive.


What do you think?
So what do you think about the carbon tax and this analysis? Does the spread of emotion represent what you are hearing? I’d love to hear your thoughts so @tweet me or leave a comment.
Disclaimer: This analysis is far from scientific. The last 1500 tweets are not representative of any significant sample size. They’re just the limit of twitter’s search API. The word list is subjective, and not a comprehensive way of performing sentiment analysis.
However, it is interesting and does give you a picture of how people are talking about the carbon tax on twitter.
