Pity poor Brazil. Not only did their team get the thumping of a lifetime on July 8th when Germany beat them 7-1, but football (soccer) fans broke every record in the book by tweeting about it.
With 36.5 million tweets seen during the match, sentiment monitors were lit up worldwide. As you might expect, the negative value for Brazil was about the same as the positive value for Germany. But what does sentiment analysis tell us?
Here at Software AG we built a World Cup Sentiment Analysis tool for anyone to enjoy while watching the football. We monitored tweets over a moving window of 30-40 minutes and scored them from positive to neutral to negative.
Twitter might just be the engine by which the mood of the planet can be measured, but it is by nature a lagging indicator.
For example, when Portugal's Christiano Ronaldo left training early one day before the team's June 22 USA match, sentiment plummeted because fans worried about an old knee injury. When the team's officials said he was fit to play, sentiment around Portugal's team rebalanced to positive.
Figuring it takes a person a little while to think of a tweet, type it and post it, we wanted to monitor as many tweets as possible for making the sentiment analysis decision. We grabbed the tweets from Twitter's public feed and dropped them into our analysis engine. The processing of the tweets is completed in under a millisecond; which means results are posted well within a second of the original tweet making its appearance.
But they are still after the fact (even if a lot of fun). So how can we use Twitter -- or other social media sentiment -- for commercial purposes?
Twitter mining is becoming the next big thing in algorithmic trading; with sentiment analysis being used to try to qualify and quantify the emotional chatter around a particular market. It then gauges whether the feelings for a particular stock or commodity are negative or positive, and uses the information for making trading decisions.
A study by the University of Manchester and Indiana University in 2010 concluded that the number of 'emotional words' on Twitter could be used to predict daily moves in the Dow Jones Industrial Average. A change in emotions expressed online would be followed between two and six days later by a move in the index, the researchers said, and this information let them predict its movements with 87.6 percent accuracy.
Another study, this one at Pace University in 2011, found that social media could predict the ups and downs of stock prices for three global brands, Starbucks, Coca-Cola, and Nike.
A U.K. hedge fund, Derwent Capital, liked the idea so much it opened an algorithmic hedge fund in 2012 that made trades based on Twitter sentiment. It soon closed, but reportedly returned 1.86 percent, beating the overall market as well as the average hedge fund.
The question is, can markets be predicted using sentiment algorithms? I think you could use a Twitter algo to get a sentiment reading on particular topics, whether it is revolutions or how people feel about the economy.
The World Cup, though, may be a different matter. You can get some interesting insights about the global consciousness surrounding a particular match, but sentiment analysis will not predict the outcome of the game. But if you could feed the sentiment analysis into another system that was set up with parameters to predict the outcome, you could be onto a winner.
Twitter sentiment analysis could be the next Paul -- the psychic octopus that made several accurate predictions in the 2010 World Cup. Paul would choose his food from two identical boxes decorated in the team flags of the upcoming matches.
Sadly, Paul died a few months later. But perhaps his legacy lives on in a real-time predictive sentiment engine somewhere.