Delivered as a project for the Dutch National Police. We publish about this work sparingly: the Police decides what we may share.
The Dutch police hold a great deal of data, but it sits in systems each built for its own purpose. A landscape like that does not grow towards coherence on its own. It grows apart.
The question
The question was not which new data should be added. The question was what becomes visible when existing sources are linked: does that yield information that is in none of the separate sources?
What we did
Twentynext carried out that analysis. Which combinations of datasets genuinely produce new, relevant information, and which mainly produce spurious correlations.
The work took place within Q-LAB, the Police innovation programme where new ways of working and technology are explored through experiments, user research and prototypes. A reference is available on request.
Why the answer is not a system
The result is not a platform, but a substantiated answer to the question of where linking pays off. There is little glory in that answer, and it saves the most money: you do not build what turns out to be noise.
For an organisation considering opening up its sources, this is the cheapest step you can take. And the only one that can justify the next.
About this article. We check facts and sources carefully, but cannot guarantee that they are accurate or complete. See our disclaimer.
