Using social media to predict crimes

Although most people do not broadcast in advance their intention to engage in criminal activity, researchers in the US have discovered that the use of Twitter could help predict crime. Matthew Gerber, assistant professor of systems and engineering information at the University of Virginia, who has been researching statistical crime prediction methods, explained: “My initial hypothesis was that there would be no correlation between Twitter use and crime. After all, people don’t share with the world that they intend to or have just committed a crime. “What they do share are things like social events or outings that could lead to criminal activity.” Professor Gerber focused his research on Twitter rather than other social media platforms because of its “openness and the fact that anyone can access GPS (global positioning system)-tagged tweets generated in a given area”. He collected more than 1.5 million public tweets tagged with Chicago-area GPS coordinates spanning January to March of 2013, as well as crime records covering the same period and geographic area. After dividing and mapping out tweets and crime records onto a grid and identifying common topics of discussion – such as sports, restaurants and entertainment – appearing in tweets, Professor Gerber combined conclusions from this analysis with older forecasting models to predict crimes over the following month. The result of his combined method was more precise, accurately predicting 19 out of 25 crime types. “Some cities that utilise such methods as a basis for resource allocation have seen dramatic decreases in crime,” said Professor Gerber, who co-directs the university’s Predictive Technology Laboratory, which uses data to create predictive models with the goal of promoting better decision-making in policing. As for the “causal connection” between tweets and crimes, he admits his method cannot answer that. Even so, it is gaining attention from police departments across the US, including Chicago and New York, and could further assist police in resource allocation, deciding where and when to deploy officers. Estimates put social media membership at approximately 2.5 billion non-unique users, and the data produced by these users has been collated to predict elections, movie revenues and even the epicentre of earthquakes. Research by Cardiff University has also shown that Twitter-monitoring systems can detect disturbances far quicker than police reports. An analysis of data taken from the London riots in 2011 showed that computer systems could automatically scan through Twitter and detect serious incidents, such as shops being broken in to and cars being set alight, before they were reported to the Metropolitan Police Service, in some cases by more than an hour. The systems could also discern information about where the riots were rumoured to take place and where groups of youths were gathering.