One of the key problems facing our clients today is gaging the emotional response to their brand or campaign. Within social data there are currently automated ways of judging the emotional context of the data, generally assigning a positive, negative or neutral sentiment. However, an inherent problem with this approach is that human language and behaviour is difficult to quantify in such simplistic terms. How many of us would now use the word “sick” now as a positive reaction?

MEC has taken on this challenge by taking the core concept of assigning text into “good” or “bad” emotions one step further. From a combination of the latest psychological theories of emotion, Corpus Linguistics and advances in Natural Language Processing we have created a bespoke tool that can split out these emotions and contextualise the deeper emotions of the consumer.

Introducing… The Emotion Wheel!


With this interactive tool we have looked at how the emotional state of the nation on Twitter has evolved for David Cameron and Ed Miliband in the run up to the General Election.

Just press and drag the selector on the graph to any date you want. Looking back to the beginning of March, you can use this tool to look at emotional changes over your selected time period and see how these two leaders compare to each other.  Scroll to the right to see how last night’s TV debate panned out.  Click on a theme to see example tweets.

Check out The Emotion Wheel – HERE! We’ll be updating it every day next week so keep checking back for updates!

If you’d like to chat to the developers Emily Sanne, John Clarvis & Tom Kendle from our fantastic A&I team, please contact them through the @mecuk Twitter handle!