Fighting crime using statistics

There is an interesting article in today's online edition of the Guardian newspaper ( about the use of statistics by police departments around the world. There is even a new term to describe this "new" approach to policing - Crush. It stands for "Criminal Reduction Utilising Statistical History" which means police forces are predictive analytics to help them fight crime.

When you think about this (bearing in mind the recent revelations about the data collection and analysis operations of the US government) all law enforcement agencies are required to consider how, when and where to deploy their resources to ensure maximum "efficiency". While the report in the Guardian focuses on the police department in Memphis, Tennessee it does mention in passing that two UK police forces who are also using Crush. I would like to add another example of this sophisticated approach to crime fighting. 

The Italian Department of Public Security of the Ministry of the Interior  has, for some time, been using Oracle Spatial and Oracle’s data mining features to help track crime, understand how crime patterns change over time and determine the most efficient way to use their resources.

The agency created an advanced data warehouse containing detailed information about criminal events. Each record covered all the aspects of each crime by including data points such as details about the victim, location information, date, time, type of crime etc. These data points allowed crime teams to evaluate existing criminal activities and patterns to help predict future criminal activities in terms of type of crime and crucially the likely location. Oracle’s in-database analytics and support for a wide variety of data types were used to deliver efficient management, integration and analysis of all the different types data recorded in the crime reports.

Oracle’s in-database Spatial and Advanced Analytics were used to enrich the basic data that was collected to generate both geospatial data points and additional statistical data points to help evaluate the relevance and confidence of the predictions.

The agency used an Oracle’s Business Intelligence tools to provide cartographic views of the geospatial information alongside the standard reporting views (tabular and graphical), which helped the crime teams get a more complete picture of the various patterns of crimes across their regions. The enriched data warehouse was able to support three levels of analysis relating to geospatial analysis, classification and grouping of types of crime and a real-time research feature.

Already today, a large number of crime teams are leveraging this new system and it has significantly increased their ability to analyze criminal activity and has reduced the time taken to respond to queries. The data warehouse is being further extended to track the progress of reported crimes with the objective being to incorporate this additional information in to the predictive model that helps plan how law enforcement resources are deployed.

This new data warehouse has helped this agency to produce predictive and proactive analysis of criminal activity. It has helped them manage their resources more effectively by increasing patrols in higher risk locations and placing offices directly in critical areas. In the future the system will allow teams to analyze specific dangerous patterns of activity using typical characteristics of a crime such as the time slot, location and the specific geospatial features of the area such as the proximity of points of interest.

If you want more information about the scenario outlined above then read this customer press release from last November (2012):

PDF Image 04 copy

There is a very good technical presentation from Iconsulting (an excellent Oracle partner that has a huge amount of experience with our spatial and data mining features) available here ( that explains how they implemented the crime dashboard for the Department of Public Security of the Ministry of the Interior.

This is a great example of Oracle's unique analytical capabilities: in-database analytical mashups. The layering of spatial analytics over data mining discoveries and then using the analytical capabilities of each feature to drive further, deeper analysis is a core strength of our database. "Take the analysis to the data" and you can open up a whole new world of discoveries. 


Technorati Tags: , , , , , ,


Popular posts from this blog

PDF version - Big Data Warehousing Must See Guide for Oracle OpenWorld 2017

Review of Big Data Warehousing at OpenWorld 2017 - Now Available

MATCH_RECOGNIZE and predicates - everything you need to know