Police forces hold masses of data about offenders, their local areas and the people they serve. Yet they often fail to make full use of it and there remains huge untapped potential for solving crimes and protecting people. Looking at that data and combining it in new ways, however, can generate insights that drive real improvements, quickly. This is what we call dark data analytics, and we’ve seen it drive huge value across the public and private sectors.
So, when we took on some police officers as interns through Police Now, we decided to work with them to understand how dark data could transform policing.
Police Now trains outstanding graduates to become police officers through a comprehensive two-year training scheme.
As part of the leadership programme, some of the graduates have a two-week internship with us. We give them specialist training in areas such as problem-solving and presentation skills, and challenge them with projects that bring together their expertise with ours to shine new light on age-old policing challenges. Dark data was one such project.
We asked the police interns to map the masses of data their force captures against its uses. We then applied our experience of dark data analytics and the interns’ knowledge of outstanding questions in policing to highlight opportunities for re-using the data to find actionable insights.
The key to the success of the project was the interns’ hunches on how to tackle neighbourhood policing (drug dealing, rough sleeping, penalty notices etc.). The interns hypothesised that organised criminals use predictable routes, community resolutions are inefficient, and Community Protection Notices are an ineffective way to deal with rough sleepers committing anti-social behaviour.
Dark data would let them test those hypotheses, leading to evidence-based decisions that add real value for the public. So, we helped the interns plan how they would take a dark data approach to these theories given the mapped data.
There are two key principles police forces need to adopt for successful dark data analysis.
The first is to start with the big hunches teams have always wanted to prove or disprove, or the gut decisions they want evidence for. By starting with the hypotheses and looking for the evidence, forces can focus resources on solving priority issues first.
The second is to not worry about analysing ‘big data’ or ‘small data’. Focusing efforts on huge, perfect datasets won’t necessarily find new insights. It’s not the size or format of the dataset that matters but the technique and logic used to answer the key questions.
While dark data can’t address all the pressures faced by policing, if deployed carefully with the right support, it can play a valuable role in managing scarce resources and helping to keep the public safe.