Big DataPredictive Analytics
Predictive Policing to Combat Terrorism

As we continue to offer tribute to family and friends affected by the recent act of terror, it’s also a time that we ought to reflect and move into inertia in our commitment to finding a lasting solution to the terror bug that has bitten us. It is a matter not just the preserve of the law enforcement, but indeed a collective responsibility of us citizens, our leaders, the media and experts at large.
There are two approaches in solving problems, one is in hindsight, reference to what has already happened often termed as epistemic and the other is foresight, looking ahead before the occurrence and preparing before it happens. The later has been mostly probabilistic and aleatoric in nature. But with the advent of technology and data, it’s becoming the industry norm to deploy preventative mechanism.
For instance after years and series of devastating earthquakes, the Japanese in 2011 embarked on a project dubbed NAMASKE, an early earthquake warning system that has over the years proven its worth as its able to alert authorities and citizens to a high degree of accuracy of a looming earthquake and necessary precautionary measures taken to avert catastrophic outcomes.
Terrorism to us is what earthquakes are to the Japanese atleast going by the trail of destruction its and the very nature of the surprising executions. We therefore need to invest in Predictive systems that can enable to send early warning signs of an impending attacks and nipping them in the bud when it happens.
Nobel Peace Prize Behavioral economist Daniel Kahneman identified one of the more common cognitive biases that humans are prone to “the representativeness bias“ When looking at historical evidence in which two things happen together: there are three possibilities Coincidence – things do happen together Direct Causation – the two things are related, so that one triggers the other Indirect Causation – the two things are each related to a third thing, that trigger each other. Overcoming the cognitive biases is the first step towards achieving a desirable outcome.
Innovative technology, once unleashed, rarely retreats. Law enforcement has always seemed to be one step behind criminals; predictive policing and pre-crime technology offer the potential to reverse that, or at least reduce the length of the step. However, citizen involvement is the essential counterweight to the growth of policing using pre-crime technology, with community leaders becoming the cornerstone to its reasonable and fair implementation.
We often see in the media reports of people eating grass as a result of committing a crime, Citizens it appears, when they take matters into their own hands, resort to psychic powers for intervention. Data, not psychic energy, drives today’s pre-crime technology. It is an attempt to apply a public health approach to crime. Just as epidemiological patterns reveal environmental toxins that can increase health risks (like getting cancer), criminal patterns can increase life risks (like getting shot).
Predictive policing requires sifting through data to identify both key risk factors and the conditions under which crimes are likely to result. Law enforcement already uses statistics to determine which roads and neighborhoods to patrol more frequently, but modern predictive policing systems takes this to a whole new level of scope and precision. One such system that rides on our human behavior of interacting is the Social Network Analysis.
A crime detection system in which a social network analysis is combined with hierarchical clustering is capable of destabilizing terrorist networks. A typical terrorism ring has 7 categories Organizers, individuals who constitute the core and control activity, Insulators – single individuals and groups which isolate the core of the organization from infiltration. Communicators – single individuals who control the flow of information between two nodes of the network. Guardians– focused on security of the network and minimization of vulnerability to external attack or infiltration. They control who is recruited and assess loyalty to the network. Extenders– deal with the extension of the network by recruiting new members and merging other networks. They encourage individuals (e.g. lawyers, policemen, politicians) from other networks to cooperate. Monitors are individuals who are responsible for effectiveness of the network. They report weaknesses to the organizers who can take remedial steps, and are responsible for improvement of the network functioning. Crossovers are individuals recruited to the network, but they still also operate in different areas (legal, governmental, financial, or commercial institutions).
Social Network Data Analysis helps Police have knowledge of the structure of connections between the roles. This knowledge allows them to verify whether these roles are correctly assigned and indicates which entities have to be monitored since they may fulfill important roles in an organization. To analyze the structure of connections between the roles, several algorithms can be implemented.