A University of Wolverhampton academic has played a leading role in a project to create the world’s first social media driven light show at the EDF Energy London Eye for the Olympics.
Mike Thelwall, Professor of Information Science at the University’s School of Technology, was asked to lend his expertise to the project to create the stunning spectacle.
Professor Thelwall is one of the leading experts worldwide in infometrics- the science of measuring information to assess the importance of a set of documents, and Mike specialises in the newest area – measuring the value of web pages.
As part of the Energy of the Nation project commissioned by EDF Energy, an official sponsor of London 2012, Olympic related tweets have been tracked since May.
Every tweet is analysed for its sentiment, whether it’s positive, negative or neutral. These are then filtered through a computer program, which converts them into a light show using the London Eye.
The colours will reflect the mood of people tweeting such as purple for negative, yellow for positive and green for neutral.
The shows use 24 hours of data and transforms it into a 24 minutes event which will run every night during the Olympics and Paralympics, starting tonight at 9pm.
Mike has developed a specific related dictionary of 2,750 terms using a system he developed called SentiStrength, which can determine the positivity or negativity of tweets.
He said: “The system basically scores every tweet according to a host of keywords and assigns it an overall sentiment. It all happens in a millionth of a second.
“The sort of work I am involved is normally hidden away in papers and offices so to be involved in something where you can see your work on a global scale, 137 metres high in the sky is fantastic. I’ve been telling everyone about it.
“Once the Olympics get under way I think you’ll see a surge of positivity. We’ve had a bit of negativity in the last week or so around security issues but once the games start and hopefully Team GB start bringing some medals back then we’ll see plenty of supportive tweets.”
Mike’s system is used by some companies to measure how people are talking or responding to their products.
He and is research team are also developing it so computers and even computer generated characters in virtual worlds can respond to human emotion.
For more information please contact James Allen in the Media Relations Office on 01902 322003