Browse Prior Art Database

Context-Aware Crime Scene Correlation

IP.com Disclosure Number: IPCOM000241448D
Publication Date: 2015-Apr-29
Document File: 3 page(s) / 166K

Publishing Venue

The IP.com Prior Art Database

Abstract

Rather than spending time in (or once released from) prison, many offenders are outfitted with electronic monitoring (EM) devices that indicate their current location and in some instances their current activity. These devices, usually attached to the ankle or other appendage, broadcast global positioning system (GPS) information as well as determine which offenders are in or near designated areas that are prohibited (designated as zones). This paper describes a method for identifying potential offender activity to a crime using the data collected by an electronic monitoring (EM) system. This activity is commonly referred to as crime scene correlation and typically returns all monitored offenders within a certain radius of the crime scene around the time that the crime occurred as potential suspects. The proposed method attempts to distinguish between offenders who were near the crime scene, but are unlikely to be potential suspects and those who are by adding a level of contextual awareness using additional information about monitored offenders such as their historical mobility patterns, criminal records, and their actions around the time of the crime.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 45% of the total text.

Page 01 of 3

Context-Aware Crime Scene Correlation

Abstract

Rather than spending time in (or once released from) prison, many offenders are outfitted with electronic monitoring (EM) devices that indicate their current location and in some instances their current activity. These devices, usually attached to the ankle or other appendage, broadcast global positioning system (GPS) information as well as determine which offenders are in or near designated areas that are prohibited (designated as zones). This paper describes a method for identifying potential offender activity to a crime using the data collected by an electronic monitoring (EM) system. This activity is commonly referred to as crime scene correlation and typically returns all monitored offenders within a certain radius of the crime scene around the time that the crime occurred as potential suspects. The proposed method attempts to distinguish between offenders who were near the crime scene, but are unlikely to be potential suspects and those who are by adding a level of contextual awareness using additional information about monitored offenders such as their historical mobility patterns, criminal records, and their actions around the time of the crime.

Introduction

A current problem with crime scene correlation is that it lacks contextual awareness. Specifically, the correlation engine returns a list of monitored individuals near the crime scene around the time of the crime with the length of time each individual was near the crime scene as well as each individual's closest distance to the crime scene, the time spent at rest, and the average speed of the offender while near the crime scene.

Some offenders are more likely to be suspects than others, depending on their historical behavior and their activities around the time that the crime occurred. The proposed context-aware crime scene correlation utilizes information about the state of the individual at the time of the crime, including whether the individual deviated from his or her usual movement patterns and the activity that the individual was engaged in (such as walking or driving), to distinguish between individuals likely to be suspects and individuals unlikely to be suspects.

The primary function of the correlation engine is to compare the locations and times of reported crimes to the locations and times of all offenders on EM devices. If any offender wearing an EM device is within the vicinity of a crime, their details will be included on a crime summary report.

Implementation

The context-aware crime scene correlation engine utilizes both long-term (historical behavior patterns) and short-term (around the time of a crime) information regarding an offender. Short- term information includes proximity to a crime scene, the amount of time spent near a crime scene, and what the offender was likely doing around the time of the reported crime. Long-term information includes the offender's criminal record as well as his or her hist...