Jacob Aron, reporter
Anyone who fondly remembers the heady days in early July 2012 when the discovery of the Higgs boson was hotly anticipated, and eventually announced, can now relive the thrilling experience thanks to an analysis of Higgs-related traffic on Twitter.
The traffic - amounting to more than 1 million tweets - provides a neat reflection of real-world excitement, starting with rumours of the elusive particle, and eventually erupting into a buzz of Higgsteria with global reach. The data might even help marketers predict how news about their products will spread on Twitter.
This video shows how the rate of tweets changed per hour over the key period, in various locations around the world.
The discovery in July of the long-sought Higgs boson, or at least something very much like it, was easily the biggest science story of 2012. Manlio De Domenico of the University of Birmingham, UK, and colleagues, who study the relationships between social and geographic networks, saw it as a unique opportunity to gather data.
Rumour watch
Normally such Twitter data sets are gathered retrospectively, but as a former physicist, De Domenico had an advantage. "I was aware something big about the Higgs was being presented, so I proposed to monitor the progress of the rumours."
Using software to monitor for the words "lhc", "cern", "boson" and "higgs", the group began collecting tweets on 1 July, when rumours of a particle discovery were already beginning and anticipation was mounting. The researchers continued their monitoring beyond the particle announcement at CERN, near Geneva, Switzerland, on 4 July, right up until the 7th.
The rate of tweets per hour on 1 July was just 36. This grew in the following days and spiked on 4 July when it reached a peak of about 36,000 an hour (see below).
To help make sense of their data, the researchers split the tweets into four time periods. The first covered the initial Higgs rumours, the second included the final release of Higgs data from the already-defunct Tevatron collider in Batavia, Illinois, which once hoped to beat the LHC to the Higgs discovery, the third covered further rumours and the fourth spanned the actual announcement of the Higgs and the aftermath.
The number of users joining the Higgs conversation changed dramatically over these four periods. In the graph below, you can see that although it looks as if there is just a small jump following CERN's announcement, the graph uses a logarithmic scale, which means that only 10 per cent of users who tweeted about the Higgs were doing so before the announcement.
Analysing Twitter users' locations revealed how the Higgs conversation changed over the seven monitored days. Initially, the conversation seemed localised, with consecutive Higgs tweets most likely to be sent by users living 20 kilometres apart. However, by the time of the announcement, consecutive tweets were equally likely to come from anywhere in the world.
Unsurprisingly, the official CERN Twitter account got the most retweets during the monitored period. Second place was less predictable, however, going to Colin Eberhardt, a software consultant who still has relatively few followers.
Eberhardt managed to strike a nerve with one of his tweets - winning the prize for the single most-shared tweet. It read: "Possibly the biggest scientific discovery of our time, the #Higgs Boson, announced in glorious MS Comic Sans Font", making reference to the odd choice of font used in the presentation from ATLAS, one of the two LHC experiments that discovered the Higgs.
No hard feelings
The New Scientist account @newscientist was beaten to third place by just one retweet by University of Manchester physicist and TV presenter Brian Cox.
As well as providing a neat anatomy of a historical event, the work helps us understand how other topics spread on Twitter, say De Domenico and his team. They used the data to create a mathematical model that uses the number of newly active Higgs tweeters at any given moment to predict the number that will emerge at future points in time.
Such information could be useful for people who want to keep a message spreading, such as marketers. "If you want to keep interest for a certain product going, you might want to estimate where and when to tweet," he says.
In some ways, it seems the announcement of a new particle isn't very different from the launch of a new product.
Reference: arxiv.org/abs/1301.2952
