I want to believe I can listen to Social Networks to understand sentiment and I can't. Its all #NickCleggsFault
Its 22nd April 2010. The British election campaign is in full swing. It probably really hit the accelerator pedal this time last week. That was when, for the first time ever, the leaders of the three main political parties performed a debate on national television. The outcome of that debate was a huge momentum swing for the Liberal Democratic party, who, until that time had been listed as the also rans probably more so in this campaign than those of the recent past. Nick Clegg's (the Leader of the Liberal Democrats) popularity and that of his party hit new highs overnight and little has happened to reverse that trend.
Tonight is the second televised election and in the lead up to this event there has been a concerted attempt to attack Nick Clegg's image and tarnish his reputation.
What, you might ask has that to do with Social Networks?
There are a number of companies who would have you believe that it is possible to monitor public opinion by simply listening to the tweets, texts and general buzz of the social networks. That software could evaulate this constant stream of data and provide clear insight into the opinions of the public. This would be a kind of Utopia for many organisations who are constantly looking to understand the opinions of its customers, prospects and the general public. Currently they have to depend on more manual research engaging Market Research departments and companies in direct contact with individuals.
So is it possible to understand whether the public is pro or anti Nick Clegg as he enters the second debate?
The above image from TweetStats shows that over the last 11 hours Nick Clegg is the fourth most tweeted term (with neither of his opponents in the top ten) while #NickCleggsFault is the second most tweeted term (more of that in a moment).
This second image from Trendistic shows that for a few hours tweets regarding Nick Clegg's rose dramatically. I personally became aware of this phenomena at 2:00pm when @BoagWorld a leading Web Designer in Britain posted the following:
This was the first in a small flurry of tweets by @BoagWorld all with the term #NickCleggsFault included. So what is this hashtag. Well someone in the twitterverse decided that Nick Clegg was having a hard time with the latest round of media attacks and that they should attack media back on his behalf. They started sending tweets out stating issues that they were having and blaming NickClegg. here is a little example taken from Twitter's search:
The question is: Any analytics available today regarding social networks like twitter seem to conveniently stop at the concept of tweet counting. There are some what try to show influence and other simple measures. There do not seem to be any that want to indicate whether there is positive and/or negative feedback and #NickCleggsfault is probably why. Reading these tweets could give a text analyser the impression that there is a lot of negative thoughts regarding Nick Clegg. Effectively it will miss the irony within them where they are not Blaming Nick at all just showing how silly it is to be negative about Nick Clegg.
Will it ever be possible to measure Social Networks effectively in anything but a human orientated manual process. Yes it may be possible to spot high flying comments and terms and bring them to someone's attention but there seems little reason to believe that the "translation" of these terms will be able to fall to any non-human being.
