Crime tracking algorithm works in real time

Posted by Kate Taylor

Israeli researchers have developed an algorithm that can be used by security organizations to track the likely location of criminals on-the-fly.

The Tel Aviv University team says its system can process new pieces of information instantly, plugging it in to existing data.

It's not dissimilar to the algorithms used by Amazon, for example, to generate purchasing suggestions based on books, music or products the user has already purchased or browsed.

"If the object is moving, modelling and eventually catching it is mathematically complex," says professor Irad Ben-Gal.

The algorithm reduces pieces of information such as phone calls, emails or credit card interactions to a set of random variables for analysis.

In a single phone call, for example, variable might include the recipient of the call, its length and the location of the caller. Once all this is known, the algorithm not only assesses patterns of crime to predict future movements, but also creates a probability map displaying possible locations for the person or group being investigated.

The map is divided into zones where the subject - who could be a criminal, a terrorist organization or a drug dealing ring - is likely operating. Each zone is given a rating showing the likelihood that the subject's there.

Although refining the programming of the original algorithm can take a few hours, says Ben-Gal, each subsequent piece of new information can be processed in milliseconds, with the analysis available instantly.

Details of the algorithms will be published soon in Quality Technology & Quantitative Management.