PRECOBS

from Wikipedia, the free encyclopedia

PRECOBS ( Pre C rime Obs ervation S ystem, seldom also: precobs or precobs ) is a software for crime prognosis. The forecast software is developed by the Oberhausen Institute for Pattern-Based Forecasting Technology (IfmPt) and marketed worldwide.

Goal setting

With PRECOBS, using crime data from the recent past, forecasts for a defined “district” are created for a police authority and used for operational measures (arrest at the crime scene) and for crime prevention . So- called predictive policing (foresighted police work) should be able to make predictions for certain offenses (e.g. break-ins, car crimes, robbery, arson). For this purpose, police data such as perpetrator profiles , focus of offenses, local conditions and empirical findings from the offense fields are evaluated. The aim is to predict crime concentrations (near repeats) in terms of time and space. In this context, a near repeat is understood to mean the occurrence of two offenses from one offense field within 72 hours in a limited geographical area .

use

PRECOBS is already being used by seven domestic and foreign police authorities in the field of criminal geography, for example in Zurich and Aargau since 2013 . In 2014, Bavaria was the first German state to use this software. PRECOBS is currently used in Munich and Nuremberg , among others . From October 2015 to April 2016 , the police in Stuttgart and Karlsruhe tested the program for an initial pilot phase in order to counter the number of break-ins, which has been increasing for years. This was followed by another one-year test with a further developed version of the forecasting software from August 2017.

The Solothurn canton police in Switzerland decided not to use it again after examining the system in 2017. The reasons given were relatively high costs and no clearly identified benefits.

The Freiburg Max Planck Institute for Foreign and International Criminal Law examined the use of software in a detailed study.

criticism

Critics fear, among other things, that the previously anonymous data collection could later be supplemented by personal data. Due to the high number of unreported cases, it is also unclear how much crime can be fought with such software. Since searches are only carried out with police-registered offenses and patterns, there is a risk that the software will narrow the view of certain locations.

Web links

See also

Individual evidence

  1. ^ Predictive Policing made in Germany . In: ifmpt.de . Retrieved January 14, 2016.
  2. ↑ Till-R. Stoldt: Police start criminal hunt with forecast software . In: welt.de . October 19, 2014. Retrieved January 14, 2016.
  3. ↑ Frank Christiansen: With "Precops" the police know where there will be a break-in next . In: augsburger-allgemeine.de . January 8, 2015. Accessed January 14, 2016.
  4. ↑ Near Repeat Prediction . In: ifmpt.de . Archived from the original on April 30, 2015. Retrieved May 1, 2015.
  5. ↑ Warning, you are at risk of a break-in . In: tagesanzeiger.ch . September 4, 2015. Accessed January 14, 2016.
  6. ↑ Jannis Brühl, Florian Fuchs: Police software for predicting crimes - Wanted: intruders of the future . In: sueddeutsche.de . September 12, 2014. Accessed March 30, 2016.
  7. ↑ Cathérine Simon: "Precob": Police want to predict break-ins using software . In: dpa.de . November 8, 2014. Retrieved January 14, 2016.
  8. ↑ Christine Bilger: New software for the police in Stuttgart The digital friend and helper . In: stuttgarter-zeitung.de . September 16, 2015. Retrieved January 14, 2016.
  9. ↑ Precobs predicts burglary locations . In: derwesten.de . September 5, 2015. Accessed January 14, 2016.
  10. ↑ Rainer Wehaus and Nils Mayer: Software should stop intruders . In: stuttgarter-nachrichten.de . July 30, 2017. Accessed March 30, 2018.
  11. ↑ Sven Altermatt: No clear benefit: Police in Solothurn do without burglary software . In: solothurnerzeitung.ch . December 12, 2017. Accessed March 30, 2018.
  12. ↑ Predictive Policing: Evaluation of the Baden-Württemberg pilot project P4 . In: mpicc.de . Retrieved March 30, 2018.
  13. ↑ Sarah Heuberger: Predictive Policing: Software is not a silver bullet against break-ins . In: wired.de . October 12, 2017. Accessed March 30, 2018.
  14. ↑ Jörg Thoma: PRECOBS: Berlin wants to predict break-ins with software . In: golem.de . December 2, 2014. Retrieved January 14, 2016.
  15. ↑ Kai Biermann: Nobody has yet proven that data mining helps the police . In: zeit.de . March 29, 2015. Retrieved January 14, 2016.