What is the system tryping to predict?

Risk Terrain Modeling tries to both predict and explain crime: Predicting where crime is more likely to occur, and explaining which terrain features most contribute to that risk.

What data are the system’s predictions based on?

RTM’s algorithms select and weight factors that are geographically related to crime incidents. RTM “takes crime data for a specific locale along with other data about the physical environment and forecasts where new crime incidents are likely to emerge and cluster.” RTM does not rely solely on past crimes to make valid forecasts.

What approach does the system use to make its predictions? How is the data analyzed?

RTM starts with the belief that specific physical features of a landscape, and the ways people interact with them, can be used to compute a probability of criminal behavior occurring nearby in the future. According to the academics who developed this approach, “RTM is by all intents-and-purposes a diagnostic method. With a diagnosis of the attractors of criminal behavior or other hazardous outcomes, you can make very accurate forecasts.”

One claimed advantage of RTM is that, while other hotspot approaches tell you where crime is clustering, that doesn’t answer the question of why it’s clustering. “While there are social, situational, political, cultural, and other factors related to the variety of crime outcomes, there is also a spatial component … RTM advances this by providing the spatial diagnosis.”

What do users of the system see or do?

RTM’s outputs can be tabular or visual. Here are some examples:

Tabular output for Robbery in Kansas City, MO highlights the spatial factors that are significant attractors and, by exclusion, those that aren’t:

Here's what their maps can look like:

The two Chicago maps show: “the micro-level places and police beats, respectively, that are high risk for Battery/Assault/Ambush against Chicago Police Officers, where the grey-scaled map of Kansas City shows the places considered to be the highest risk for robbery Kansas city.

How accurate are the predictions? How is accuracy defined and measured?

RTM claims to provide a final model that ‘paints a picture’ of places where criminal behavior is “statistically most likely to occur.” Though “[w]e are not at the point where we can precisely predict specific crime events by particular offenders at specific moments in time … but, with RTM, police predict, with a certain level of confidence, the most likely places where crimes will emerge and cluster.”

Further, RTM claims that it is a “sustainable technique because past crime data are not always needed to continue to make valid forecasts.”

Which departments are using the system?

RTM is reportedly now used by researchers or practitioners in over 30 countries across 6 continents.

  • Kansas City PD
  • Newark PD
  • Atlantic City PD
  • New York City PD
  • New Haven PD
  • Jersey City PD
  • City of Fayetteville PD

How are departments using the system?

TK.

results matching ""

    No results matching ""