Computer Scientists Use Twitter to Predict UK General Election Result

by  Lee Page,  University of Warwick

Computer scientists from the University of Warwick used Twitter to predict the outcome of the U.K. general election. They are working in collaboration with partners in the Department of Journalism at City University London and the Information Technologies Institute (ITI-CERTH, Greece).

The team has developed an algorithm that harvests political tweets, and incorporating sentiment conveyed in tweets was one of its key features. The user-generated content is aggregated and put into conventional polling reports to produce a daily prediction of voting share. “We then put all this information into our forecasting model, along with the parties’ share of the vote as measured by opinion polls,” says Warwick researcher Adam Tsakalidis.

The team says the approach will provide key insights into how public opinion is developing and what factors might be influencing any changes in support. The researchers believe their forecasts could be more accurate than traditional opinion methods.

City University’s Professor Steve Schifferes added: “Social media is now central to how people consume election news. This ground-breaking research will give us deeper insights into how people are developing their own understanding of the election that goes beyond the traditional news media’s agenda, and could enable us to spot early trends which could affect the overall outcome.”

Tested during the Greek election in January, the model achieved better results than all of the most recent polls leading up to the vote and three exit polls once the ballots closed.

“We are particularly interested in automatically identifying the sentiment expressed towards specific politicians or parties and topics such as immigration,” Tsakalidis says. “This will help us obtain more accurate predictions as well as better understanding of the reasons behind public support or discontent.” Article

DCL: Similar techniques have been employed in SE Asia using cell phone traffic and rumor reports in local presses (the equivalent of Twitter at the time) to detect the emergence of epidemics.  This proved highly successful even when national governments attempted to suppress the news. Early detection of infectious diseases is a major issue in today’s air travel world. There’s a lot of event processing going on in these detection systems.