A group of computer scientists from the UAE, US and UK have won a seemingly impossible worldwide challenge: to track down five ‘suspects’ of a jewel heist in five different cities within 12 hours. Their win redefines the limits of technology-mediated social mobilization and rapid information gathering.
The ‘Tag Challenge’, took place last month – a not-for-profit, independent event, aiming to “determine whether and how social media can be used to accomplish a realistic, time-sensitive, international law enforcement goal.” With a $5,000-reward, it constituted a simulated law enforcement search in five cities, namely Washington DC, New York City, London, Stockholm and Bratislava in Slovakia. Sponsored by the US State Department and the US Embassy in Prague, the challenge was created by a group of graduate students from six countries.
On the morning of 31 March, the organizers posted on the website a “mug shot” of each suspect at 8:00am local time. Using this photo alone, teams competed to be the first to successfully locate and photograph all ‘volunteer suspects’ and submit verifiable photographs to the contest organizers before 8:00pm local time.
The enormity of the challenge meant that no single person or group of friends could tackle it on their own. Instead, winning was expected to rely on the ability to assemble a very large, ad-hoc team of spotters. A group of computer scientists provided the tool to do just that.
The winning team, dubbed CrowdScanner was built by computer scientists from the Masdar Institute in Abu Dhabi, in collaboration with researchers at the University of Southampton and University of California San Diego (UCSD). They located three targets in New York City, Washington DC, and Bratislava in Slovakia within the stipulated time. Two other teams were subsequently able to locate one suspect each, in New York and Bratislava. Suspects in Stockholm and London remained at large.
The team leader, Professor Iyad Rahwan of Masdar Institute in Abu Dhabi, said: “What is most fascinating about our success, in my opinion, is the fact that none of us resides in any of the five cities. We all coordinated everything from behind our computer screens. This, to me, is what is special about the Internet. Our next step is to try to reconstruct exactly how we won, and what happened on the day of the challenge, and to learn something about what makes social media work in such amazing ways.”
Dr Rahwan added: “Back in the 1960s, psychologist Stanley Milgram conducted the legendary “six degrees of separation” experiment, which redefined our conception of social distance and taught us that everyone is only 6 steps away from everyone else on earth. The Tag Challenge is similarly redefining the limits of technology-mediated social mobilization and rapid information gathering. This has implications ranging from disaster response, to finding missing children, and much more.”
Dr Victor Naroditskiy of the Agents, Interaction and Complexity group at the University of Southampton commented: “The TAG Challenge is a chance to study how information propagates through social networks and what it takes for a message to go viral. Our analysis of the data from the challenge is aimed at discerning general patterns that will help explain the successes and failures of social media in recent events.”
Members of the AIC group, led by Professor Nick Jennings in ECS-Electronics and Computer Science, have been working together on the use of crowdsourcing in information-gathering over the last year, particularly as regards theoretical work on game theory and incentives. The global group included research contacts around the world, resulting in a balanced mix of practical and game-theoretic expertise.
Professor Jennings leads the ORCHID project, funded by EPSRC and industrial partners, which aims to establish the science needed to understand, build and apply human-agent collectives (HACs), with crowdsourcing one of the major technologies under investigation and study.