Crime Analysts Research Articles

Partially Supervised Spatiotemporal Clustering for Burglary Crime Series Identification

SummaryStatistical clustering of criminal events can be used by crime analysts to create lists of potential suspects for an unsolved crime, to identify groups of crimes that may have been committed by the same individuals or group of individuals, for offender profiling and for predicting future events. We propose a Bayesian model-based clustering approach for criminal events. Our approach is semisupervised, because the offender is known for a subset of the events, and utilizes spatiotemporal crime locations as well as crime features describing the offender's modus operandi. The hierarchical model naturally handles complex features that are often seen in crime data, including missing data, interval-censored event times and a mix of discrete and continuous variables. In addition, our Bayesian model produces posterior clustering probabilities which allow analysts to act on model output only as warranted. We illustrate the approach by using a large data set of burglaries in 2009–2010 in Baltimore County, Maryland.

Predictive Policing: The Future of Law Enforcement in the Trinidad and Tobago Police Service (TTPS)

The Trinidad and Tobago Police Service (TTPS) is currently faced with large volumes of criminal data that continues to grow daily and which are required to be processed and transformed into useful information and where data mining can greatly improve crime analysis and aid in preventing and reducing crime. Currently, crime analysts attached to the analytical department of the TTPS are required to unravel the complexities in data to assist operational personnel in arresting offenders and also in directing crime prevention strategies. With the current volume of crime being committed and the awareness of modern criminals, this is becoming a very daunting task. The ability to analyze huge volumes of data with its inherent complexities without the use of computational support puts a strain on human resources. Because of the speed and advances in the field of data mining within recent years, independent studies on its impact on policing are only now getting on the way. It is particularly important in this respect to examine the benefits which the TTPS can derive through a careful implementation of this technology. The infamous events of July, 1990 in Trinidad and Tobago heralded the need for predictive policing and exacerbated concerns about national security by the local law enforcement agency. Accurately and efficiently analyzing the organization’s ever growing volume of crime data is a major challenge facing the TTPS. This paper presents a case for implementing data mining (knowledge discovery in databases -KDD) within the TTPS as a tool for predictive analytics of crime data. It is hoped that this technology will provide decision makers with intelligence from the crime data to inform their strategic planning. It discusses the challenges of implementing data mining with special discussion of key issues relating to data integrity and the information technology (IT) infrastructure required to support data mining. It concludes by suggesting the internal information technology (IT) infrastructural changes needed to facilitate its implementation in the TTPS.

Joint Utility of Event-Dependent and Environmental Crime Analysis Techniques for Violent Crime Forecasting

Violent crime incidents occurring in Irvington, New Jersey, in 2007 and 2008 are used to assess the joint analytical capabilities of point pattern analysis, hotspot mapping, near-repeat analysis, and risk terrain modeling. One approach to crime analysis suggests that the best way to predict future crime occurrence is to use past behavior, such as actual incidents or collections of incidents, as indicators of future behavior. An alternative approach is to consider the environment in which crimes occur and identify features of the landscape that would be conducive to crime. Thanks to advances in geographic information system technology and federally funded (free) software applications such as CrimeStat III or the Near Repeat Calculator, these methods have recently been made more accessible to “average” users. This study explores the information products that each method offers for the purposes of place-based violent crime forecasting and resource allocation. Findings help to answer questions about where, when, and why violent crimes occur in a jurisdiction. Ways in which event-dependent and environmental crime analysis techniques can be utilized as complementary instruments in a crime analyst’s tool kit are discussed in detail.

Differentiating Genuine and False Rape Allegations: A Model to Aid Rape Investigations

The aim of this study was to compare the victimology, offence demographics and offence behaviours of genuine and false allegations of rape in order to develop a model that could be used to aid in differentiating between genuine and false rape allegations. Eighty false allegations of rape were compared with 160 genuine allegations of rape, obtained from the Serious Crime Analysis Section (SCAS). Data were taken directly from complainant statements and 62 variables were dichotomously coded. Chi-square, odds ratios and Mann-Whitney U tests were initially completed and then the more significant variables were entered into a backwards stepwise logistic regression. The resultant logistic regression model classified 91.7% of the sample correctly. The model was validated using an additional 12 cases, with an 83% success rate. This study found that a number of factors successfully discriminated between genuine and false rape allegations. It thus offers a tool that might aid police investigations, crime analysts and behavioural investigative advisors in the veracity assessment of rape allegations.

An analysis of crime hot spots using GPS tracking data of children and agent-based simulation modeling

Although the problem of a lack of an adequate denominator of crime rates has been widely recognized, relatively few solutions have been found. This paper proposes a sophisticated method of analyzing crime hot spots using global positioning system (GPS) tracking data and agent-based simulation modeling. The goal of the proposed analysis technique is to estimate a population at risk across time and space at the micro-level, such as streets and locations (e.g. parks and playgrounds). In order to create valid estimates of a population at risk and to illustrate how the estimated population at risk can be used in crime analysis, this paper presents a series of three sub-studies. First, GPS tracking data were obtained for two weeks for 80 children in the second and fifth grades from one school located in a suburb of Tokyo, Japan. GIS software was used to analyze the GPS data in order to identify streets and places that the children frequented. Second, agent-based simulation modeling was developed to extend the proposed analysis method to school districts that had not participated in the GPS sub-study. In the process of developing a simulation algorithm, the validity of simulation modeling was checked by comparing the simulation results with the GPS tracking data. Third, victimization surveys with maps were conducted in six schools (including the school that participated in the GPS sub-study) that identified locations of crime incidents. Victimization survey results were overlaid with the analysis results of GPS data and agent-based simulation modeling in order to examine crime hot spots. The proposed technique of agent-based modeling produces an accurate denominator of crime rates at the micro-level which can then be used to analyze crime hot spots properly. In particular, this paper discusses how crime analysts can identify different types of hot spots, such as crime generators and crime attractors, by examining the number of crime incidents and the magnitude of crime rate. This paper argues intervention strategies need to be tailored for each type of crime hot spots, as these crime hot spots develop through different mechanisms. Finally, this paper discusses limitations of the proposed technique and the direction of future research.

Translating Environmental Criminology Theory into Crime Analysis Practice

In theory, the concepts of environmental criminology are introduced into the work of police departments by crime analysts. In ideal applications, the analysts use their knowledge of place, opportunity, and the situational nature of crime when assessing criminogenic factors to guide the development of police responses. A test of this theoretical process shows us that there is a disconnection in the transition of knowledge from theory to application. We use a survey of crime analysts to identify sources of this disconnect between the two fields as well as discuss potential remedies to the problem. Increased interaction between the two fields and recognition of the valuable contributions of each will lead to better analysis and problem-solving by both environmental criminologists and crime analysts, and ultimately the reduction and prevention of crime.

The effective analyst: a study of what makes an effective crime and intelligence analyst

The current study aimed to identify the skills and abilities required by an analyst to be recognised as effective. Thirty subject matter experts (SMEs) were engaged using the Repertory Grid Technique (RGT) to explore the specific skills and abilities of crime and intelligence analysts that result in them being deemed effective. Semi-structured interviews and the Critical Incident Technique (CIT) were conducted to strengthen these findings. Three clusters of variables were identified that indicate an effective analyst; they would have skills and abilities that could contribute to the development and dissemination of an analytical product, they would have an attitude that included being productive, seeking out work, having a high-level commitment and pride and having a ‘can do’ attitude. Some physical and personal attributes, like gender and age, were included that characterise effectiveness. The findings are discussed in relation to earlier descriptions of the skills needed to be an analyst. The findings illustrate a change from the analyst being seen as a technical specialist to a growing understanding of the analyst as part of a support structure for decision-makers. Implications of the findings for recruitment, training and development are discussed.

The Missing Link of Crime Analysis: A Systematic Approach to Testing Competing Hypotheses

Crime analysts have traditionally received little guidance from academic researchers in key tasks in the analysis process, specifically the testing of multiple hypotheses and evaluating evidence in a scientific fashion. This article attempts to fill this gap by outlining a method (the Analysis of Competing Hypotheses) of systematically analysing multiple explanations for crime problems. The method is systematic, avoids many cognitive errors common in analysis, and is explicit. It is argued that the implementation of this approach makes analytic products audit-able, the reasoning underpinning them transparent, and provides intelligence managers a rational professional development tool for individual analysts.

A Preliminary Examination of Crime Analysts' Views and Experiences of Comparative Case Analysis

Comparative Case Analysis (CCA), typically conducted by crime analysts, uses crime scene behaviours to try to identify series of crimes committed by the same offender. Accurate identification of series of offences allows the police to pool resources and evidence, thereby boosting the potential to identify and apprehend the offender. This paper discusses the results of a survey of a sample of crime analysts working in two UK police forces about their views and experiences of CCA. The results focus on how CCA is conducted, what evidence and information is considered, and how useful CCA is to criminal investigation. Suggestions for how CCA might be developed further are also included.

Data mining and crime analysis

AbstractAn essential component of criminal investigation involves the interrogation of large databases of information held by police and other criminal justice agencies. Data mining and decision support systems have an important role to play in assisting human inference in this forensic domain that creates one of the most challenging decision‐making environments. Technologies range widely and include social network analysis, geographical information systems, and data mining technologies for clustering crimes, finding links between crime and profiling offenders, identifying criminal networks, matching crimes, generating suspects, and predicting criminal activity. This paper does not intend to cover the gamut of techniques available to the investigator of crime as this has been presented elsewhere (Oatley GC, Ewart BW, Zeleznikow J. Decision support systems for police: lessons from the application of data mining techniques to ‘soft’ forensic evidence. Artif Intell Law 2006, 14:35–100). Rather, the objective is to highlight issues of implementation and interpretation of the techniques available to the crime analyst. To this end, the authors draw from their experiences of working with real‐world crime databases (Oatley GC, Belem B, Fernandes K, Hoggarth E, Holland B, Lewis C, Meier P, Morgan K, Santhanam J, et al. The gang gun‐crime problem—solutions from social network theory, epidemiology, cellular automata, Bayesian networks and spatial statistics. Accepted: book chapter for IEEE publication Computational Forensics; 2008; Oatley GC, McGarry K, Ewart BW. Offender network metrics. WSEAS Trans Inf Sci Appl 2006, 3:2440–2448; Oatley GC, Ewart BW. Crimes analysis software: pins in maps, clustering and Bayes net prediction. Expert Syst Appl 2003, 25:569–588), involving gun and gang crime, fraud, terrorism, burglary, and retail crime. © 2011 John Wiley & Sons, Inc. WIREs Data Mining Knowl Discov 2011 1 147‐153 DOI: 10.1002/widm.6This article is categorized under: Algorithmic Development > Spatial and Temporal Data Mining Application Areas > Science and Technology Application Areas > Society and Culture Fundamental Concepts of Data and Knowledge > Human Centricity and User Interaction

Risk Clusters, Hotspots, and Spatial Intelligence: Risk Terrain Modeling as an Algorithm for Police Resource Allocation Strategies

The study reported here follows the suggestion by Caplan et al. (Justice Q, 2010) that risk terrain modeling (RTM) be developed by doing more work to elaborate, operationalize, and test variables that would provide added value to its application in police operations. Building on the ideas presented by Caplan et al., we address three important issues related to RTM that sets it apart from current approaches to spatial crime analysis. First, we address the selection criteria used in determining which risk layers to include in risk terrain models. Second, we compare the “best model” risk terrain derived from our analysis to the traditional hotspot density mapping technique by considering both the statistical power and overall usefulness of each approach. Third, we test for “risk clusters” in risk terrain maps to determine how they can be used to target police resources in a way that improves upon the current practice of using density maps of past crime in determining future locations of crime occurrence. This paper concludes with an in depth exploration of how one might develop strategies for incorporating risk terrains into police decision-making. RTM can be developed to the point where it may be more readily adopted by police crime analysts and enable police to be more effectively proactive and identify areas with the greatest probability of becoming locations for crime in the future. The targeting of police interventions that emerges would be based on a sound understanding of geographic attributes and qualities of space that connect to crime outcomes and would not be the result of identifying individuals from specific groups or characteristics of people as likely candidates for crime, a tactic that has led police agencies to be accused of profiling. In addition, place-based interventions may offer a more efficient method of impacting crime than efforts focused on individuals.

(Looking) Back to the Future: Using Space—Time Patterns to Better Predict the Location of Street Crime

Crime analysts attempt to identify regularities in police recorded crime data with a central view of disrupting the patterns found. One common method for doing so is hotspot mapping, focusing attention on spatial clustering as a route to crime reduction (Chainey & Ratcliffe, 2005; Clarke & Eck, 2003). Despite the widespread use of this analytical technique, evaluation tools to assess its ability to accurately predict spatial patterns have only recently become available to practitioners (Chainey, Tompson, & Uhlig, 2008). Crucially, none has examined this issue from a spatio-temporal standpoint. Given that the organisational nature of policing agencies is shift based, it is common-sensical to understand crime problems at this temporal sensitivity, so there is an opportunity for resources to be deployed swiftly in a manner that optimises prevention and detection. This paper tests whether hotspot forecasts can be enhanced when time-of-day information is incorporated into the analysis. Using street crime data, and employing an evaluative tool called the Predictive Accuracy Index (PAI), we found that the predictive accuracy can be enhanced for particular temporal shifts, and this is primarily influenced by the degree of spatial clustering present. Interestingly, when hotspots shrank (in comparison with the all-day hotspots), they became more concentrated, and subsequently more predictable. This is meaningful in practice; for if crime is more predictable during specific time-frames, then response resources can be used intelligently to reduce victimisation.

Dimensionality Analysis as a Computerized Tool for Strategic Planning in Policing and Security

Multidimensional scaling is a useful addition to the toolkit of the crime or security analysts, made possible by the advent of desktop computing power. Three uses of this graphical statistical tool are as follows: (a) forming policy initiatives and making resource allocations; (b) analyzing terrorism threats and dynamics; and (c) performing structural analysis of performance data. Dimensionality analysis is compared with more commonly encountered techniques such as factor analysis, cluster analysis, and correspondence analysis, and their relative merits discussed. Data used to illustrate dimensionality analysis are drawn from the crime statistics on US cities, terrorism incidents in Israel, and performance indicators for American police departments, but the methods presented are broadly applicable to a wide variety of policing and security issues and the same principles apply to other forms of data.

Effects of prior interpretation on situation assessment in crime analysis

AbstractA main threat to objective information processing in crime investigation teams is the tendency to focus on one particular interpretation only. To prevent such tunnel vision or ‘groupthink’, an investigation team can call in a crime analyst, and ask him or her to give a fresh and independent account of the evidence at hand. However, before they examine the case, crime analysts are often already aware of the scenario currently favoured by the team. In our experiment, we investigated whether such prior knowledge can jeopardise the independence of the analyst's advice. Thirty‐eight professional crime analysts were asked to generate causal scenarios for two different cases and to indicate how the team should continue their investigation. Before beginning their analysis, half of the crime analysts received a realistic prior interpretation, such as might have been constructed by an investigation team. The results show that when given a prior interpretation, both experienced and inexperienced analysts considered the scenario suggested therein as more likely, and made recommendations for further investigation accordingly. We explain these findings by suggesting that analysts temporarily adopted the perspective of the investigation team, and that such temporary commitment by itself increased confidence in the hypothesis at hand (Koehler, 1991). This research supports previous research on the impact of prior theory on judgement, and extends it to an important real world domain where mistakes can have serious consequences. We recommend that in cases where the crime analyst is asked to give an objective assessment, he or she should not be informed about the interpretation of the investigation team until after the analysis has been conducted. Copyright © 2007 John Wiley & Sons, Ltd.

The Integration of Crime Analysis Into Law Enforcement Agencies

The purpose of this article is exploratory and it seeks to depict how crime analysts perceive patrol officers' and other sworn police officers' attitudes about the crime analysis function and its “organizational fit” as well as their own attitudes about sworn personnel. These issues were explored with data collected through an online survey of crime/intelligence analysts. The main findings are that most analysts felt that management was fairly supportive of their work, but that their perceptions of what patrol officers thought of them was middle of the road, at best. This is contrary to the much more positive attitudes they expressed about patrol officers. The results raise concerns about whether analysts are being accepted within the police culture and whether patrol officers understand what analysts do, which in turn leads to additional concerns about whether the skills and training of analysts are being used to their fullest capacities.

Visualising space and time in crime patterns: A comparison of methods

Previous research exploring space–time patterns has focused on the relative merits and drawbacks of the effectiveness of static maps vis-à-vis interactive dynamic visualisation techniques. In particular, they have tended to concentrate on the role of animation in interpretation of patterns and the understanding of underlying factors influencing such patterns. The aim of this paper is to broaden this debate out to consider the effectiveness of a wider range of visualisation techniques in permitting an understanding of spatio-temporal trends. The merits of three visualisation techniques, (map animation, the comap and the isosurface) are evaluated on their ability to assist in the exploration of space–time patterns of crime disturbance data. We conclude that each technique has some merit for crime analysts charged with studying such trends but that further research is needed to apply the techniques to other sources of crime data (and to other sectors such as health) to permit a comprehensive evaluation of their respective strengths and limitations as exploratory visualisation tools.

Forensic drug Intelligence: An important tool in law enforcement

Organised criminality is a great concern for national/international security. The demonstration of complex crimes is increasingly dependant on knowledge distributed within law-enforcement agencies and scientific disciplines. This separation of knowledge creates difficulties in reconstructing and prosecuting such crimes. Basic interdisciplinary research in drug intelligence combined with crime analysis, forensic intelligence, and traditional law enforcement investigation is leading to important advances in crime investigation support. Laboratory results constitute one highly dependable source of information that is both reliable and testable. Their operational use can support investigation and even provide undetected connections or organisation of structure. The foremost difficulties encountered by drug analysts are not principally of a chemical or analytical nature, but methodologies to extract parameters or features that are deemed to be crucial for handling and contextualising drug profiling data. An organised memory has been developed in order to provide accurate, timely, useful and meaningful information for linking spatially and temporally distinct events on a national and international level (including cross-border phenomena). Literature has already pointed out that forensic case data are amenable for use in an intelligence perspective if data and knowledge of specialised actors are appropriately organised, shared and processed. As a particular form of forensic case data, the authors’ research focuses on parameters obtained through the systematic physical and chemical profiling of samples of illicit drugs. The procedure is used to infer and characterise links between samples that originate from the same and different seizures. The discussion will not, however, focus on how samples are actually analysed and compared as substantial literature on this topic already exists. Rather, attention is primarily drawn to an active and close collaboration between magistrates, forensic scientists, law enforcement investigators and crime analysts from different institutions with the aim of generating, using and validating relevant profiling case data as integral part of investigative and crime analysis processes. Original advances are highlighted through experiences from criminal investigations of offences related to the unlawful importation, exportation, supply and possession of illicit drugs.

An associate constraint network approach to extract multi-lingual information for crime analysis

International crime and terrorism have drawn increasing attention in recent years. Retrieving relevant information from criminal records and suspect communications is important in combating international crime and terrorism. However, most of this information is written in languages other than English and is stored in various locations. Information sharing between countries therefore presents the challenge of cross-lingual semantic interoperability. In this work, we propose a new approach – the associate constraint network – to generate a cross-lingual concept space from a parallel corpus, and benchmark it with a previously developed technique, the Hopfield network. The associate constraint network is a constraint programming based algorithm, and the problem of generating the cross-lingual concept space is formulated as a constraint satisfaction problem. Nodes and arcs in an associate constraint network represent extracted terms from parallel corpora and their associations. Constraints are defined for the nodes in the associate constraint network, and node consistency and network satisfaction are also defined. Backmarking is developed to search for a feasible solution. Our experimental results show that the associate constraint network outperforms the Hopfield network in precision, recall and efficiency. The cross-lingual concept space that is generated with this method can assist crime analysts to determine the relevance of criminals, crimes, locations and activities in multiple languages, which is information that is not available in traditional thesauri and dictionaries.

Geocoding crime and a first estimate of a minimum acceptable hit rate

Spatial crime analysis relies not only on accurate geocoding but also the achievement of a high level of geocoding success. Geocoding is the task of converting locations, such as the addresses of burglary victims, into grid coordinates and is a task performed regularly by many crime analysts. Data sources include police offence and incident databases where the quality of geographical references can vary. The reality of dealing with this real world data means that achieving a completely successful geocoding process is rare and few crime analysts can get a hit rate (the percentage measure of success) of 100%. This paper seeks the answer to a seemingly simple question: what is an ‘acceptable’ minimum geocoding hit rate for crime data? This paper uses a number of different crime patterns and Monte Carlo simulation to replicate a declining geocoding hit rate to answer this question. Reduced crime rates of mapped points, aggregated to census boundaries, are compared for a statistically significant difference. The result indicates 85% as a first estimate of a minimum reliable geocoding rate, and this result is applicable to many address-based, point pattern datasets beyond the crime arena.

Data association methods with applications to law enforcement

Associating records in a large database that are related but not exact matches has importance in a variety of applications. In law enforcement, this task enables crime analysts to associate incidents possibly resulting from the same individual or group of individuals. In practice, most crime analysts perform this task manually by searching through incident reports looking for similarities. This paper describes automated approaches to data association. We report tests showing that our data association methods significantly reduced the time required by manual methods with accuracy comparable to experienced crime analysts. In comparison to analysis using the structured query language (SQL), our methods were both faster and more accurate.