Detecting serial residential burglaries using clustering
Introduction
Studies suggest that a large proportion of crimes are committed by a minority of offenders, e.g. in the USA, researchers suggest that 5% of offenders are involved in 30% of the convictions (Tonkin, Woodhams, Bull, Bond, & Palmer, 2011). Law enforcement agencies, consequently, are required to detect series of crime, or linked crimes. A series can be defined as multiple offences committed by a serial offender. A serial offender can be defined as someone who has committed two or more crimes of the same type (Woodhams, Hollin, & Bull, 2010). It is suggested by law enforcement in Sweden that, similarly to the international findings, a large proportion of the residential burglaries are committed by professional criminals that travel across large areas of Sweden. Simultaneously, according to the Swedish National Council for Crime Prevention, law enforcement agencies solved approximately three to five percent of the 21,300 reported residential burglaries in 2012.
The detection of linked crimes is helpful to law enforcement for several reasons. Firstly, the aggregation of information from crime scenes increases the amount of available evidence. Secondly, the joint investigation of multiple crimes enables a more efficient use of law enforcement resources (Woodhams et al., 2010).
Law enforcement needs to handle a large amount of reported crimes, and the detection of series of crimes are often carried out manually. A decision support system that enables law enforcement to decrease the amount of cases when reviewing crimes would increase resource efficiency.
Forensic evidence, e.g. DNA, and fingerprints, has been used to detect linked crimes (Bennell and Canter, 2002, Tonkin et al., 2011). The availability of forensic evidence is, however, limited (Tonkin et al., 2011). In the absence of forensic evidence, behavioural information can be used as an alternative data source (Bennell & Canter, 2002). A criminal committing a series of crimes has been found to have a high intra-crime behavioural similarity (Woodhams et al., 2010). Similarly, behavioural consistency tends to be lower between criminals in similar situations (Woodhams et al., 2010).
This article presents a new decision support system (DSS) that can be used to systematically collect burglary data and to perform visualisations, analyses, and interpretations of the collected data. The article evaluates a key component of the DSS: the use of clustering techniques to group burglaries based on different definitions of similarity between burglaries, described in Fig. 1. Clustering has been used to group data according to similarity between data points, or to find communities in the data. Clustering residential burglaries based on different similarity aspects would potentially allow law enforcement to find series whilst reviewing a smaller amount of residential burglaries, i.e. used as a case selection DSS. Consequently, the use of this DSS would allow law enforcement agencies to save resources, whilst providing individual investigators with increased support. The clustering is performed using the cut clustering algorithm (Flake, Tarjan, & Tsioutsiouliklis, 2004).
The purpose of this study is twofold. First, a DSS for collecting, managing and analysing residential burglary information is presented. Secondly, the potential of minimum cut based graph clustering of crimes is investigated to reduce the amount of crimes to review to detect series of residential burglaries. The impact of different edge representations and edge removal criteria on cluster quality and accuracy is investigated. Clustering quality is measured using the modularity index and accuracy is evaluated by applying the rand index.
The data comprises residential burglary reports gathered from southern Sweden and the Stockholm area.
The remainder of this work is organized as follows: Section 2 presents a DSS for residential burglary analysis. In Section 3, the related work is reviewed. Section 4 then describes the minimum cut clustering algorithm. In Sections 5 and 6, the methodology and experimental procedure is described. The results of the experiments are presented in Section 7 and analysed in Section 8. Conclusions and future work is presented in Section 9.
Section snippets
Decision support system for residential burglary analysis
Since 2011, researchers from Blekinge Institute of Technology collaborate with law enforcement officers and analysts from the Blekinge county police as well as four additional county police authorities from southern Sweden. The aim is to develop Information and Communication Technology (ICT) solutions for law enforcement. The scope is currently limited to solutions that target residential burglaries. The strategies, tactics, and overall organisational structure of the police vary between
Related work
The problem of linking reported crimes has mostly been investigated from a psychological or criminological perspective. The research has focused on crimes conducted that can be considered violent, e.g. sexual offences, rapes, homicides, and different types of burglaries, including violent burglaries (Bennell and Canter, 2002, Bennell et al., 2010a, Bennell et al., 2010b, Bennell et al., 2010c, Markson et al., 2010, Woodhams et al., 2010).
The research conducted suggests that behavioural
Cut clustering algorithm
The cut clustering algorithm is a graph-based clustering algorithm based on minimum cut tree algorithms to cluster the input data (Flake et al., 2004). The input data used is an undirected graph where the edges between nodes could represent a similarity or distance measure.
The algorithm can be described as follows: an artificial node is added to the existing graph and connected to all nodes in the graph with the edge value . A tree is created from the graph using the minimum cut tree algorithm
Data collection
The data consist of residential burglary incident reports collected in a systematic way by law enforcement officers over a period of six months. The incident reports are collected through a checkbox-based form, providing a common base of data collected. The form used consists of eleven sections and 107 checkboxes. In addition to the checkboxes, information about time, date and geographical position (longitude, latitude and street address) of the reported incident is also gathered. If required,
Experiment design
The following two aspects of residential burglary clustering are investigated. First, the impact of different similarity indices as edge representations and of different edge removal criteria on the quality of the clusters produced. Second, the performance with which the minimum cut algorithm is able to group residential burglaries without splitting series of crimes.
Experiment 1
According to the modularity cluster validation measure (see Table 1(a)), the 1st-Quantile has the worst performance of all the different edge removal criteria. Similarly, the edge Jaccard Goods and Temporal proximity representations are performing worse than other representations. The performances of these edge representations indicate that Jaccard Goods and Temporal proximity are unsuitable for representing differences between crime cases. The goods available in the form are a few general
Analysis
The results of the experiments are analysed using an ANOVA test to detect if there exist any statistically significant difference between the variables. Fisher’s LSD test is used to detect between which variables statistical significant differences exist.
Conclusion
In this article a DSS for managing and analysing systematically gathered residential burglary reports have been presented. The DSS allows law enforcement to easily search and compare residential burglary reports. The DSS contains, among other modules, an analytical framework. The use of clustering to group residential burglaries in the DSS has been investigated, using several similarity criteria.
While results of the modularity cluster validation measurement indicate that the separation between
References (30)
- et al.
Linking commercial burglaries by modus operandi: Tests using regression and ROC analysis
Science & Justice: Journal of the Forensic Science Society
(2002) - et al.
The impact of data degradation and sample size on the performance of two similarity coefficients used in behavioural linkage analysis
Forensic Science International
(2010) Community detection in graphs
Physics Reports
(2010)Graph clustering
Computer Science Review
(2007)- et al.
Linkage analysis in cases of serial burglary: Comparing the performance of university students, police professionals, and a logistic regression model
Psychology, Crime & Law
(2010) - et al.
Between a ROC and a hard place: A method for linking serial burglaries bymodus operandi
Journal of Investigative Psychology and Offender Profiling
(2005) - et al.
Addressing problems with traditional crime linking methods using receiver operating characteristic analysis
Legal and Criminological Psychology
(2010) - Brandes, U., Delling, D., Gaertler, M., Görke, R., Hoefer, M. Nikoloski, Z., et al. (2007). On finding graph...
- et al.
GIS and crime mapping
(2005) - Cohen, J., Rodrigues, L. A., Silva, F., Carmo, R., Guedes, A. L. P., & Duarte, E. P. (2011). Parallel implementations...
Graph clustering and minimum cut trees
Internet Mathematics
Dynamic graph clustering using minimum-cut trees
Cited by (28)
A deep learning framework for predicting burglaries based on multiple contextual factors
2022, Expert Systems with ApplicationsCitation Excerpt :We also use criminals’ source address distributions for each region, which indicate how many criminals are living in each region. In the field of computational criminology, criminals’ MO, have been shown to be effective for solving crimes (Bennell & Canter, 2002; Borg et al., 2014). Knowledge of a criminal’s MO is required for performing crime linkage (Markson et al., 2010), i.e., identifying a set of crimes that were committed by the same criminal, e.g., routinely breaking in through a window, always stealing identity cards, etc.
Smart meter data classification using optimized random forest algorithm
2022, ISA TransactionsAn anticrime information support system design: Application of K-means-VMD-BiGRU in the city of Chicago
2022, Information and ManagementCitation Excerpt :Related to the crime pattern theory, many of the existing studies use a space-centric paradigm, and the focus of research is to predict where crime events occur. Given that crime event occurrences are not equally distributed across a city, most studies tend to identify high-crime-intensity areas and forecast crime hotspot areas where hotspot recognition is an application of spatial data clustering [21]. Corcoran et al. [22] presented an approach for crime incident forecasting, which focused on geographical areas of concern that exceed traditional policing boundaries.
Thresholds learning of three-way decisions in pairwise crime linkage
2022, Applied Soft ComputingCitation Excerpt :Many crimes are committed by repeat offenders [1].
Static or dynamic? Characterize and forecast the evolution of urban crime distribution
2022, Expert Systems with ApplicationsA novel random forest approach for imbalance problem in crime linkage
2020, Knowledge-Based SystemsCitation Excerpt :Serial crimes refer to two or more crimes committed by the same criminal or the same group of criminals (()). Serial crimes account for a large proportion of the total number of crimes (()). Therefore, in some countries, governments require the police to focus on investigating serial offenders who are the minority of offenders committing the majority of crimes (()).