Validating distance decay through agent based modeling (original) (raw)

Strengthening Theoretical Testing in Criminology Using Agent-based Modeling

The Journal of research in crime and delinquency, 2014

The Journal of Research in Crime and Delinquency (JRCD) has published important contributions to both criminological theory and associated empirical tests. In this article, we consider some of the challenges associated with traditional approaches to social science research, and discuss a complementary approach that is gaining popularity-agent-based computational modeling-that may offer new opportunities to strengthen theories of crime and develop insights into phenomena of interest. Two literature reviews are completed. The aim of the first is to identify those articles published in JRCD that have been the most influential and to classify the theoretical perspectives taken. The second is intended to identify those studies that have used an agent-based model (ABM) to examine criminological theories and to identify which theories have been explored. Ecological theories of crime pattern formation have received the most attention from researchers using ABMs, but many other criminologica...

An Agent-Based Model for Public Security Strategies by Predicting Crime Patterns

IEEE Access

In recent years, statistical methods have been applied to the study of crime patterns. However, these schemes have several drawbacks that prevent accurate modeling of complex behaviors. Agent-based models (ABM) allow the modeling of human behavior by employing simple rules that consider each agent's neighborhood. In this paper, a new agent-based model is proposed to emulate crime patterns produced by the interaction of different urban actors, such as offenders (criminals), citizens, and defenders (police officers). Using this approach, the simulation results provide escape trajectories and robbery frequencies that can be used to create or improve public security strategies. Although our scheme can be generically applied, we validated the model by considering different scenarios for the case of Guadalajara, Mexico. Experimental results show that the proposed scheme creates realistic offender behaviors that efficiently predict criminal patterns and provides essential data that allow the creation and improvement of public security strategies to reduce the number of crimes. INDEX TERMS Agent-based model (ABM), crime simulations, crime patterns, routine activity, computer simulation.

Agent-based modelling as a research tool for criminological research

Crime Science, 2015

Computational modelling techniques, originating from fields like Computer Science and Artificial Intelligence, may be beneficial for criminological research. Because of their formal nature, computational models can be processed by machines that operate on them, for example for the purpose of simulation. As a consequence, these techniques may help gain insights that lacked based of purely informal theories. A well-known example of such a technique, which has become widely applied within criminology, is called agent-based modelling. Agent-based modelling (ABM) is a computational method that enables a researcher to create, analyse and experiment with models composed of agents, i.e., autonomous pieces of software that interact within a computational environment (Gilbert, 2008). In the current article this technique will be explored in depth. First, I will give a description of the technique and present the architecture of an ABM. Subsequently, I will apply the technique to a simple toy example in the context of a simulation model of the bystander effect, to demonstrate the possibilities of the approach. I will discuss some pros and cons of the approach and present related work to help appreciate the benefits of applying ABM to different criminological research questions. Hopefully, this will provide readers with the necessary knowledge to consider the use of ABM in their own research.

Implementing comprehensive offender behaviour in a realistic agent-based model of burglary.

"Explaining and modelling crime patterns is an exercise that has taxed policy-makers, criminologists, social reformers and the police ever since the first crime patterns were recorded. Crime is a particularly difficult phenomenon to model because of its inherent complexity; crime patterns are built up from a multitude of human-human and human-environment micro-interactions that ultimately lead to individual crime events. Commonly used modelling techniques, such as regression, struggle to fully account for the dynamics of the crime system. They work at aggregate scales thereby disregarding important individual-level variation and also struggle to account for the effects of different types of human behaviour. Furthermore, important concepts from environmental criminology – such as individual offender awareness spaces or heterogeneity in offender decision-making – cannot be included directly when working at a resolution above that of the individual. This research addresses the drawbacks associated with traditional mathematical crime models by building an agent-based simulation with a unique offender behavioural model. Through use of the PECS framework for modelling human behaviour, agents are endowed with needs and motives that drive their behaviour and ultimately lead to the commission of crime. As the model uses real-world environmental data, it can be used to make predictions in existing cities. The paper demonstrates that use of this framework, in combination with an agent-based model, can replicate patterns and trends that are supported by the current theoretical understanding of offending behaviour. "

Comparing Crime Prevention Strategies by Agent-Based Simulation

2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, 2009

Within the field of Criminology, an important challenge is to investigate the spatio-temporal dynamics of crime. Typical questions in this area are how the emergence of criminal hot spots can be predicted and prevented. This paper presents an agentbased simulation approach that is able to address such questions. More specifically, the approach can be used to compare different strategies for guardian movement in terms of their effectiveness. To illustrate the approach, a number of simulation experiments have been performed, and the results are discussed. † In this paper we focus explicitly on crimes that are performed on the street against random passers-by, e.g. pick-pocketing.

Agent-Based Simulation of Crime

2013 12th Mexican International Conference on Artificial Intelligence, 2013

The effects of crime are diverse and complex, ranging from psychological and physical traumas faced by crime victims, to negative impacts on the economy of a whole nation. In this paper, an agent-based crime simulation framework to analyze crime and its causes is proposed and implemented. The agent-based simulation framework models and simulates both 1) crime events as a consequence of a set of interrelated social and individual-level crime factors, and 2) crime opportunities, i.e., combinations of circumstances that enable a person to commit a crime. The selection of crime factors and design of agent models are supported by, and based on, existing criminological literature. In addition, the simulation results are validated and compared with macrolevel crime patterns reported by various criminological research efforts.

Statistical Models of Criminal Behavior: The Effects of Law Enforcement Actions

Mathematical Models and Methods in Applied Sciences, 2010

We extend an agent-based model of crime-pattern formation initiated in Short et al. by incorporating the effects of law enforcement agents. We investigate the effect that these agents have on the spatial distribution and overall level of criminal activity in a simulated urban setting. Our focus is on a two-dimensional lattice model of residential burglaries, where each site (target) is characterized by a dynamic attractiveness to burglary and where criminal and law enforcement agents are represented by random walkers. The dynamics of the criminal agents and the target-attractiveness field are, with certain modifications, as described in Short et al. Here the dynamics of enforcement agents are affected by the attractiveness field via a biasing of the walk, the detailed rules of which define a deployment strategy. We observe that law enforcement agents, if properly deployed, will in fact reduce the total amount of crime, but their relative effectiveness depends on the number of agents...

State of the Art in Agent-Based Modeling of Urban Crime: An Overview

Journal of Quantitative Criminology

Objectives Agent-based modeling (ABM) is a type of computer simulation that creates a virtual society and allows controlled experimentation. ABM has the potential to be a powerful tool for exploring criminological theory and testing the plausibility of crime prevention interventions when data are unavailable, when they would be unethical to collect, or when policy-makers need an answer quickly. This paper takes stock of the current literature to discuss the potential contributions of ABM, assess current practice, identify shortcomings that threaten the validity of findings using ABM, and to make suggestions regarding the construction and communication of future work using ABM. Methods We systematically searched major databases to find all publications using ABM to simulate urban crime patterns and coded publications to quantify the following information: (1) characteristics of the publication, the model and the agents, (2) model purpose, (3) crime type investigated, and (4) interrogation of the model via sensitivity testing and validation. Results After sifting papers according to our inclusion criteria, we identified and reviewed 45 publications. Models informed by the opportunity theory framework dominated. Most publications lacked detail sufficient to enable replication. Many did not include clear a rationale for modeling choices, parameter selection or calibration. Rarely were parameters

Agent-based simulation of episodic criminal behaviour1

Multiagent and Grid Systems, 2014

Criminal behaviour often involves a combination of physical, mental, social and environmental (multi-)agent aspects, such as neurological deviations, hormones, arousal, (non)empathy, targets and social control. This paper contributes a dynamical agent-based approach for analysis and simulation of criminal behaviour, covering the above aspects, illustrated for the case of an Intermittent Explosive Disorder. It involves dynamically generated desires and beliefs in opportunities within the social environment, both based on literature on criminal behaviour.