Algorithmic crime prevention. From abstract police to precision policing.

The growing digitisation in our society also affects policing, which tends to make use of increasingly refined algorithmic tools based on abstract technologies. But the abstraction of technology, we argue, does not necessarily entail an increase in abstraction of police work. This paper contrasts the 'abstract police' debate with an analysis of police practices that use digital technologies to achieve greater precision. While the notion of abstract police assumes that computerisation distances police officers from their community, our empirical investigation of a geo-analysis unit in a German Land Office of Criminal Investigation shows that the adoption of abstract procedures does not by itself imply a detachment from local reference and community contact. What we call contextual reference can be productively combined with the impersonality and anonymity of algorithmic procedures, leading also to more effective and focused forms of collaboration with local entities. On the basis of our empirical results, we suggest a more nuanced understanding of the digitalisation of police work. Rather than leading to a progressive estrangement from the community of reference, the use of digital techniques can enable experimentation with innovative forms of 'precision policing', particularly in the field of crime prevention.

First, do no harm: Critically revisiting contemporary approaches to child sexual abuse prevention.

Crime prevention is typically presented in a tripartite model that includes primary, secondary, and tertiary domains. Almost every criminal justice intervention constitutes tertiary prevention and occurs reactively, in the aftermath of an offence. Child sexual abuse is no exception, and prevention science has long recommended we focus our intervention efforts further upstream. Such an approach would include earlier detection and disclosure (secondary prevention), or-even better-reducing the risks of early exposure to the environmental forces which facilitate sexual abuse in the first place (primary prevention). What is missing from the field, however, is a coherent framework through which to critique the unintended consequences of our well-intentioned responses to child sexual abuse. Such consequences include secondary trauma for victim survivors and vicarious trauma for families and practitioners. In this article, we reflect on prevention from a critical perspective that centres the principle of "first, do no harm." In doing so, we introduce the notion of 'quaternary prevention' for child sexual abuse. Public health has long recognised the risks of medicalisation, overdiagnosis, and unnecessary intervention. We encourage our field to engage within a framework of quaternary prevention to consider the iatrogenic effects of many contemporary practices and to take seriously the "do no harm" principle to improve practice across all levels of prevention.

Detection dogs fighting transnational narcotraffic: performance and challenges under real customs scenario in Brazil.

Narcotic Detection Dogs (NDDs) are essential tools in the fight against drug trafficking, acting with high precision and improving efficiency at border posts. When trained efficiently, these dogs can detect a great variety of compounds, such as cocaine, marijuana and its derivatives, and synthetic drugs, among others. Most of the knowledge on canine detection processes and efficiency has been determined in experimentally controlled conditions, but narcotic seizures detected by dogs in realistic anti-drug operations have not yet been critically determined in a Country with continental dimensions such as Brazil. This study aimed to evaluate the data set concerning the performance, operations, efficiency, and success rate of NDDs used by the Brazilian Customs Authority (Aduana) in the fight against drug trafficking. Narcotic seizure rates increased in luggage and packages detected by NDDs working at border crossings from 2010 to 2020, with an estimated value of over US$ 2 billion in losses to the cocaine drug trafficking business. NDD units also increased most narcotic groups seized in the same period. The number of NDDs and anti-drug operations, and Customs Border Post (CBP) influenced the rates of drugs seized. NDDs provided an increase of 3,157 kg/animal of drugs seized for every new dog introduced into the inspection systems.

A Deeper Look at the Boy Scouts of America "Perversion" Files: Structural Factors Related to Access and Abuse.

The Boy Scouts of America's (BSA's) Ineligible Volunteer (IV) files, commonly called the "perversion files," is a unique data set allowing researchers to examine organizational characteristics that allow for child sexual victimization. Despite the uniqueness of this data set, few researchers have examined it. The researchers examined a random sample of cases from the IV files of scout leaders who molested scouts. A situational crime theoretical model was employed as had been used in studies on the Catholic Church sexual abuse scandal, proving effective. The researchers examined randomly selected from 1980 to 1999. Of initially examined 400 randomly selected cases, only 140 held retrievable data. The majority of the 140 cases did not have scouting victims, which resulted in a reduction to 48 cases. Beyond the several scouting forms, many files contained public domain information, including newspaper articles, police reports, criminal justice records, and/or records of civil litigation. Through both the scouting documents and the public domain records, the researchers identified six BSA-specific characteristics that allowed motivated adult leaders to molest their victims. These characteristics included (a) weak or inefficient incident reporting system, (b) failure to collect and review pertinent information, (c) organizationally legitimate reasons for one-on-one contact, (d) volunteer imbalance, (e) legitimate reasons for separation from protective adults, and (f) social status as a prophylactic defense. These structural characteristics are explored through descriptive statistics and specific case studies illustrating the phenomena. When the history of the IV files came to light through litigation, the BSA was forced to reconcile its past actions and develop new preventive measures. The Boy Scouts of America implemented various actions to protect the youth. Many of these protections specifically address structural characteristics.

'Led by Intelligence': A Scoping Review on the Experimental Evaluation of Intelligence-Led Policing.

Intelligence-led policing (ILP) was introduced in the 1990s as a proactive approach to policing, but to date, there is a lack of studies that have synthesized and summarized the central characteristics and insights of (quasi-)experimental studies related to ILP. This study aims to address this gap by synthesizing and characterizing the central characteristics of 38 quasi-experimental and experimental studies related to ILP. In this study, a scoping review is conducted on different quasi-experimental and experimental studies that relate to the framework of ILP. It was found that most studies within the domain of ILP focus on testing the crime reduction effects of using spatio-temporal crime intelligence to deploy police resources more efficiently and effectively. However, some studies have combined different types of crime intelligence or used solely offender-related intelligence. Several statistical-methodological challenges were also identified that should be considered when designing experimental research within the domain of ILP. Additionally, most studies focused solely on measuring crime reduction, with few focusing on secondary effects of interventions. The review concludes that future evaluation studies should consider evaluating the use of different types of crime intelligence and establish specific, objective, and realistic criteria for measuring specific performance measures such as crime disruption. Future experimental research within the domain of ILP should consider applying the 3-i model, evaluating each leg of ILP thoroughly. The limitations of the study are also discussed. This review provides valuable insights for future research and development of ILP-related approaches.

"It's a preventable type of harm": Evidence-based strategies to prevent sexual abuse in schools.

The last two decades have seen global public recognition of the scale and impact of adult-perpetrated institutional child sexual abuse. A sizeable body of knowledge about generalized safeguarding measures has since been generated to inform organizations' prevention efforts. Apparent in the extant literature, however, is a notable lack of evidence-based and context-specific prevention strategies targeting perpetration in distinct institutional environments. This absence extends to educational settings where most reported contemporary cases occur or originate. The recommendations outlined in this article contribute to this gap. Derived from empirical findings establishing the role of context-specific rather than person-specific factors, a range of prevention strategies framed by Situational Crime Prevention are proposed for secondary educational settings. These recommendations are supplemented by the unique insights of interviewed experts with specialist professional knowledge. Recommendations center around targeting features of the environment such as high-risk locations, available guardians, and intimacy-promoting situations available in educational contexts. Taken together, this robust prevention and control agenda can create conditions for safer educational environments.

Suspicious Behavior Detection with Temporal Feature Extraction and Time-Series Classification for Shoplifting Crime Prevention.

The rise in crime rates in many parts of the world, coupled with advancements in computer vision, has increased the need for automated crime detection services. To address this issue, we propose a new approach for detecting suspicious behavior as a means of preventing shoplifting. Existing methods are based on the use of convolutional neural networks that rely on extracting spatial features from pixel values. In contrast, our proposed method employs object detection based on YOLOv5 with Deep Sort to track people through a video, using the resulting bounding box coordinates as temporal features. The extracted temporal features are then modeled as a time-series classification problem. The proposed method was tested on the popular UCF Crime dataset, and benchmarked against the current state-of-the-art robust temporal feature magnitude (RTFM) method, which relies on the Inflated 3D ConvNet (I3D) preprocessing method. Our results demonstrate an impressive 8.45-fold increase in detection inference speed compared to the state-of-the-art RTFM, along with an F1 score of 92%,outperforming RTFM by 3%. Furthermore, our method achieved these results without requiring expensive data augmentation or image feature extraction.

Mimicry Deception Theory applied to sexual abuse of children.

Sexual abuse of children remains a widespread problem with well-documented, adverse consequences. Often, abuse ending is contingent on a disclosure made by the victim, but victims delay disclosure if they tell someone at all. The factors associated with (non)disclosure are complex and interrelated. In this paper, we propose a new theoretical framework (Mimicry Deception Theory; MDT) to explore various aspects of the grooming process, using a qualitative content analysis of US court appeal cases (N = 25). Specifically, we focus on how MDT components contribute to the likelihood of a CSA disclosure. MDT is made up of five components: Victim Selection, Community Integration, Complexity of Deception, Resource Extraction, and Detectability. These five components allow us to look at several characteristics of abuse in tandem and examine how they interact to impact various outcomes, such as (non)disclosure. We provide a detailed codebook for this framework, that can be used to systematically extract relevant information from large amounts of data. Through the application of this framework, we were able to identify several factors that may play a role in delayed or non-disclosure. Further, we found repeat offenders were likely to use the exact same methods of access, grooming, and remaining undetected across victims. Implications for prevention, as well as clinical interventions with perpetrators as well as victims are discussed.

(Re)proposing problem-oriented policing as a framework for identifying new and enhanced ways to prevent the abuse of at-risk adults.

It is recognized worldwide that evidence on what effectively prevents abuse of at-risk adults (including older people) is lacking. The purpose of this paper is to make an explicit, contemporary proposal as to why front-line workers and policy-makers interested in preventing the abuse of at-risk adults should adopt an opportunity-focused, problem-oriented policing framework from criminology. This paper will first position the proposal within the current context of responses to the abuse of at-risk adults. Following this, the paper will outline how the problem-oriented policing framework can develop targeted, collaborative, sustainable, and measurable prevention strategies intended to remove the opportunity for specific examples of abuse of at-risk adults.

Ranking Trust Factors Affecting Risk Perception in Illicit Drug Purchase on the Darknet: A Large-Scale Survey Study in Hungary.

The process of illicit drug trafficking on darknet markets is highly affected by various trust factors. Although the factors potentially affecting customers' risk perception can be identified based on previous research, cyber criminology has not produced empirical research ranking the importance of the specific factors. This study was designed to fill this gap by developing a tool that measures the importance of the various trust factors. To test out the measurement tool, a large-scale survey with projective situational questions was conducted among university students in Hungary. The sample (n = 5481) was compiled to include potential darknet market customers, respondents with above-average computer skills needed to access the darknet, and taking into account that university students are a group of society particularly exposed to drug consumption. The end product of this research is a trust matrix ranking the factors affecting illicit drug purchases on darknet markets. Among the factors, the survey's target group ranked reliable and undamaged delivery of goods and the reliability of vendors as the most important. The measurement tool developed in this research will facilitate further criminological research on vendor reputation. Its findings also point to the need for further research on delivery providers and predict that influencing the delivery-related risk perception of potential customers could effectively reduce demand.

Criminology and Crime Science in the Arab World.

Throughout most of the twentieth century, Western criminological thought was preoccupied by theories of modern criminology, which were concerned with the question of why individuals commit crime. Since the late 1970s, however, the Western world has experienced new reconfigurations in the crime control field, one of which is the transformation of Western criminological thought. New crime theories have emerged in which the focus has shifted from the 'criminal' to the 'criminal activity'. Despite their importance, these developments have had no impact on the criminological thought in the Arab world, which is, to this day, is dominated by Western theories of modern criminology. This paper has three purposes. First, it discusses the status quo of criminology in the Arab world. Second, it suggests plausible explanations that help understand the current status of criminology in this region. Third, it discusses the relevance of some criminological developments (notably crime theories and environmental criminology) in the Arab context. In that eventuality, it argues that the Western developments might bring insightful crime prevention lessons for the Arab world.

Spatio-temporal Crime Analysis and Forecasting on Twitter Data Using Machine Learning Algorithms.

The concept of social media began to gain popularity in the late 1990s and has played a significant role in connecting people across the globe. The constant addition of features to old social media platforms and the creation of new ones have helped amass and retain an extensive user base. Users could now share their views and provide detailed accounts of events from worldwide to reach like-minded people. This led to the popularization of blogging and brought into focus the posts of the commoner. These posts began to be verified and included in mainstream news articles bringing about a revolution in journalism. This research aims to use a social media platform, Twitter, to classify, visualize, and forecast Indian crime tweet data and provide a spatio-temporal view of crime in the country using statistical and machine learning models. The Tweepy Python module's search function and '#crime' query have been used to scrape relevant tweets under geographical constraints, followed by substring-keyword classification using 318 unique crime keywords. The Bokeh and gmaps Python modules create analytical and geospatial visualizations, respectively. Time series forecasting of crime tweet count is performed by comparing the accuracy of Long Short-Term Memory (LSTM), Auto-Regressive Integrated Moving Average (ARIMA), and Seasonal Auto-Regressivee Integrated Moving Average (SARIMA) models to determine the best model.

Cost-benefit analyses of developmental crime prevention programmes.

BACKGROUND: Children and youth who are at risk of becoming early-onset life-course-persistent offenders often slip through the cracks of other systems in society (e.g., health, education, child welfare, substance use and mental health). When they do, they impose an enormous economic burden on society. Developmental crime prevention (DCP) programmes seek to reduce these costs through evidence-based interventions that target individual child and family risk and protective factors for antisocial behaviour. AIM: This study reviewed cost-benefit analysis studies of DCP interventions to identify whether they produced monetary benefits that exceeded programme costs. METHOD: We searched the literature for studies of interventions that were evaluated using high-quality research methods (i.e., experimental or quasi-experimental designs). Key characteristics of these evaluations are summarised and benefit-to-cost ratios (BCRs) are reported. RESULTS: Eleven cost-benefit analysis (CBA) evaluations met study criteria. The programmes varied in terms of who they targeted (e.g., pregnant mothers, at-risk youth), the age of participants (e.g., adults, children, older youth), the intervention duration (e.g., 10 weeks to 4 years), and the follow-up interval (e.g., 6 months to 50 years). Ten of the 11 studies produced favourable BCRs, ranging between 1.35 and 31.77, depending on the type and scope of outcomes that were monetised. CONCLUSION: There is strong evidence in support of DCP from a cost-benefit perspective. However, given the small number of studies for analysis, more prospective longitudinal CBA evaluations are needed, in addition to greater consistency in the scope and methods that are used to monetise outcomes.

Prevalence of Physical Violence in the Medical-Forensic Approach in the Years 2015-2020 in City and Neighboring Municipalities: Perspectives from Poland-Poznan Study.

Forensic medical opinions serve the appropriate classification of a crime against health. Violence, a multifaceted phenomenon, requires forensic medical examination in the case of causing damage to health. Due to the effects caused by the perpetrator, the damage to health is divided into severe, medium, and light. This study analyzed 7689 incidents of violence from 2015-2020, taking place in the area subordinate to the Provincial Police Headquarters in Poznan, based on anonymized documentation of forensic medical examinations performed at the Department of Forensic Medicine in Poznan at the request of the Police and privately. The analysis took into account: units ordering the test, type of exposure, medical help, sex and age of the victim, places of the incident, classification and localization of injury, manner of impact, attitude of the perpetrator to the victim, profession of the victim, gender of the perpetrator, and remarks. In Poland, statistics on violence victims are underestimated, resulting from the low reporting of crimes committed to law enforcement authorities. There is a need for programs to educate the perpetrator of violence on methods of conflict resolution and programs to prevent violence, covering events taking place in public spaces.

Noncompleted Sexual Offenses: Internal States, Risks and Difficulties Related to Crime Commission through the Lens of Sexual Offenders.

The study aims to generate insights from sexual offenders on the influence of internal states and how they perceive risks of apprehension and difficulties in the context of noncompleted sexual offenses, that is when offenders initiated the offense but were stopped or discouraged either before or during sexual contact. Adult males incarcerated for sexually offending completed a self-report questionnaire. Regression models, including interaction effects, were estimated. Two interaction effects were found providing insights into which and how internal states, such as intoxication to alcohol, may influence perceived difficulties related to crime. Future research should promote the investigation of noncompleted sexual offenses, which could provide a real opportunity to generate new or complementary insights for better understanding and guiding prevention initiatives.

Community Resilience during the COVID 19 Pandemic: Experiences of Community-Based Violence Prevention and Recidivism Reduction Program Administrators.

Community-based agencies play a notable role in local violence prevention and reentry services in the United States. The COVID-19 pandemic and governmental responses to contain its spread fundamentally transformed the day-to-day lives of most individuals and the workplace. This study examined the challenges experienced and adaptations employed by community-based organizations as they navigated shelter-in-place orders and other workplace and community restrictions. Between July and September of 2020, researchers completed 16 semi-structured interviews with agency administrators of community-based organizations serving at-risk youth or formerly incarcerated persons operating in a large Midwestern city. The findings highlight several challenges faced by agency administrators as they attempted to maintain services to their clients, including having to move from largely in-person service modalities to methods of contact and communication that embraced social distancing and virtual interaction. They also actively responded to the health safety needs of their staff, clients, and community by instituting new safety protocols, like staff and client COVID-19 testing, handing out personal protection equipment and supplies, and educating community members. The findings demonstrate a high degree of community mobilization and resilience in light of a global crisis.

Community engagement, greening, and violent crime: A test of the greening hypothesis and Busy Streets.

Researchers have documented that vacant lot greening can reduce community-level crime and violence. Busy Streets Theory (BST) suggests that residents who are involved in the greening process can help to improve physical environments and build social connections that deter crime and violence. Yet few researchers have explored how community engagement in the greening process may affect crime and violence outcomes. We applied BST to test the effects of community-engaged vacant lot greening compared to vacant lots that received either professional mowing or no treatment, on the density of violent crime around study lots. Using mixed effects regression models, we analyzed trends in violent crime density over the summer months from 2016 to 2018 at 2102 street segments in Youngstown, OH. These street segments fell within 150 meters of an intervention parcel that was classified as one of three conditions: community-engaged maintenance, professional mowing, or no treatment (control). We found that street segments in areas receiving community-engaged maintenance or professional mowing experienced greater declines in violent crime density than street segments in areas receiving no treatment, and more decline occurred in the community-engaged condition compared to the professional mow condition. Our findings support BST and suggest that community-engaged greening of vacant lots in postindustrial cities with a concentrated vacancy can reduce crime and violence.

Effects of building demolitions on firearm violence in Detroit, Michigan.

Former industrial cities facing economic challenges and depopulation often experience high levels of firearm and other forms of violence. Within these cities, violent crime often clusters in neighborhoods affected by high levels of vacant and abandoned housing. This study estimates the effects of building demolition in Detroit, Michigan on the subsequent risk of violent crime using property-level data and longitudinal targeted maximum likelihood estimation. The primary outcome is violent Crime Index crimes (homicide, rape, robbery and aggravated assault). We estimate effects for this category of crimes as a whole and for the subset involving firearms. Drug and other lower-level crimes are included as secondary outcomes. We compare the risk of experiencing each crime type following building demolition in Census blocks and block groups to an estimate of the risk had there been no demolition in the 1-3 quarters prior in 2017. There were >2600 total demolitions in about 1700 blocks in 2017 in Detroit. Nearly all demolished buildings were sourced from tax foreclosures. Estimates suggest the risk for all crime types tested would have been statistically indistinguishable from the observed crime risk had demolitions in the prior 1-3 quarters of 2017 not occurred. Our results run counter to most previous research on this topic, which tends to show a protective effect of demolition on violent crime. Understanding why our results differ may provide important insights into the types of demolition programs with the greatest potential to reduce violent crime.

Knowledge Development Trajectories of Crime Prevention Domain: An Academic Study Based on Citation and Main Path Analysis.

This study performed main path analysis to explore the academic field of crime prevention. Studies were collected from the Web of Science database, and main path analysis was used to analyze the studies and identify influential authors and journals on the basis of the g-index and h-index. Cluster analysis was then performed to group studies with related themes. Wordle was used to output keywords and word clouds for each cluster, both of which were used as reference to name each cluster. Five clusters were identified, namely crime displacement control, crime prevention through environmental design, developmental crime prevention, the effects of communalism on crime prevention, and the effect of childhood sexual abuse on crime. Each cluster was analyzed, and suggestions based on the results are provided. The main purpose of crime prevention is to advance our understanding of the psychological criminal mechanisms (i.e., personal, social and environmental impacts) associated with different criminal behaviors at the intersection of law by using main path analysis.