Spatio-Temporal Analysis of Crime Incidents for Forensic Investigation

Crime analysis and mapping has been routinely employed to gather intelligence which informs security efforts and forensic investigations. Traditionally, geographic information systems in the form of third-party mapping applications are used for analysis of crime data but are often expensive and lack flexibility, transparency, or efficiency in uncovering associations and relationships in crime. Each crime incident and article of evidence within that incident has an associated spatial and temporal component which may yield significant and relevant information to the case. Wide variations exist in the techniques that departments use and commonly spatial and temporal components of crime are evaluated independently, if at all. Thus, there is a critical need to develop and implement spatio-temporal investigative strategies so police agencies can gain a foundational understanding of crime occurrence within their jurisdiction, develop strategic action for disruption and resolution of crime, conduct more informed investigations, better utilize resources, and provide an overall more effective service. The purpose of this project was to provide foundational knowledge to the investigative and security communities and demonstrate the utility of empirical spatio-temporal methods for the assessment and interpretation of crime incidents. Two software packages were developed as an open source (R) solution to expand current techniques and provide an implementable spatio-temporal methodology for crime analysis. Additionally, an actionable method for near repeat analysis was developed. Firstly, the premise of the near repeat phenomenon was evaluated across crime types and cities to discern optimal parameters for spatial and temporal bandwidths. Using these parameters, a method for identifying near repeat series was developed which draws inter-incident linkages given the spatio-temporal clustering of the incidents. Resultant crime networks and maps provide insight regarding near repeat crime incidents within the landscape of their jurisdiction for targeted investigation. Finally, a new approach to the geographic profiling problem was developed which assesses and integrates the travel environment of road networks, beliefs and assumptions formed through the course of the investigation process about the perpetrator, and information derived from the analysis of evidence. Each piece of information is evaluated in conjunction with spatio-temporal routing functions and then used to update prior beliefs about the anchor point of the perpetrator. Adopting spatio-temporal methodologies for the investigation of crime offers a new framework for forensic operations in the investigation of crime. Systematic consideration about the value and implications of the relationship between space, time, and crime was shown to provide insight regarding crime. In a forward-looking sense this work shows that the interpretation of crime within a spatio-temporal context can provide insight into crime occurrence, linkage of crime incidents, and investigations of those incidents