Identification of operational demand in law enforcement agencies

Triad influence on the detection of crime in Hong Kong

We use bootstrap data envelopment analysis, adjusting for endogeneity, to examine police efficiency in detecting crime in Hong Kong. We address the following: (i) is there a correlation between the detection of crime and triad influence? (ii) does the level of triad influence affect the efficiency in translating inputs (police strength) into outputs (crime detection)? and (iii) how can the allocation of policing resources be adjusted to improve crime detection? We find that nighty-eight percent of Hong Kong police districts in our sample were found to be inefficient in the detection of crime. Variation was found across districts regarding the detection of violent, property and other crimes. Most inefficiencies and potential improvements in the detection of crime were found in the categories violent and other crimes. We demonstrate how less efficient police districts can modify police resourcing decisions to better detect certain crime types while maintaining current levels of resourcing. Finally, we highlight how the method we outline improves efficiency estimation by adjusting for endogeneity and measuring the conditional efficiency of each district (i.e. the efficiency of crime detection taking the instrumental variables (e.g. influence of triads) into consideration). The use of frontier models to assist in evaluating policing performance can lead to improved efficiency, transparency, and accountability in law enforcement, ultimately resulting in better public safety outcomes and publicly funded resource allocation.

Innovation technology opportunity identification of civil aircraft mechanical connections based on generative topographic mapping

In order to advance civil aircraft manufacturing to higher levels, there is an urgent need to identify technological innovation opportunities to help new technology development. This paper first analyses the current state of the research field and determines the topic. It preprocesses papers and patents within the research topic to obtain a base database. Then, the database is analyzed using the LDA (Latent Dirichlet Analysis) cluster analysis method. The TF-IDF (Term Frequency-Inverse Document Frequency) algorithm processes the data to obtain critical technical words. The abstracts of patents and papers are processed to construct a binary-based vector of technical keywords. The papers and patents are visualized in a two-dimensional space technology map by generative topographic mapping (GTM) to create a technology map to identify technology blank dots. The combination of technologies characterized by each technology blank dot is obtained by GTM inverse mapping. Finally, technology opportunities with a high probability of development are identified to achieve innovation opportunity identification. It also provides countermeasures for the research institution, enterprise, sector, and industry. After research and analysis, the future in the mechanical connection technology of civil aircraft is necessary to strengthen basic technology development and improve the study of intelligence, integration, and flexibility. Technology such as sensors and lasers can improve the precision and efficiency of mechanical connections.

Adapting the influences of publishers to perform news event detection

Online news outlets have the power to influence public policy issues. To understand the opinions of the people, many government departments check online news outlets to manually detect events that interest people. This process is time-consuming. To promptly respond to public expectations, this research proposes a framework for detecting news events that may interest government departments. This article proposes a method for finding event trigger words used to represent an event. The news media can be a critical participant in ‘agenda-setting’, which means that more widely discussed news is more attractive and critical than news that is less discussed. However, few studies have considered the influence of news media publishers from the ‘agenda setting’ perspective. Therefore, this study proposes an ‘agenda setting’-based filter to establish a high-impact news event detection model. The proposed framework identifies trigger words and utilises word embedding to find news event–related words. After that, an event detection model is designed to determine the events that are attractive to government departments. The experimental results show that purity increases from 0.666 when no extraction method is used to 0.809 when the extraction method in this study is used. The overall improvement trend shows significant improvement in event detection performance.

Knowledge discovery in research on policing strategies: an overview of the past fifty years

Purpose

The insecurity generated, today in various parts of the planet, by the various conflicts that arise in the violence in large cities, has motivated the academy to research the solutions and strategies adopted by local governments in the fight against crime. The volume of data generated by several universities over the past 50 years has increased exponentially. Consequently, researchers struggle to process essential data in today’s competitive world. The aim of this study is to explore and provide an overview of the studies carried out in the field of action to combat crime in different countries.

Design/methodology/approach

The Web of Science and Scopus databases were searched for publications from January 1945 to September 3, 2020 on the topic of policing strategies in titles, abstracts and keywords. References were analyzed using the R bibliometrix package, and abstracts were analyzed using latent Dirichlet allocation (LDA) with collapsed Gibbs sampling for topics related to policing and related subjects.

Findings

As a result of the research, this paper can assert that in the last 50 years, 3,361 authors have produced 2,085 documents on the theme of policing strategy and related subjects in 58 countries. Scientific production in this area grows at a rate of 5.10 per year. The USA is the leading country in publications with 42.58%, followed by the UK with 8.39% and Canada with 4.07%. As for journals, the highlight is Policing, Policing and Society and Police Quarterly, which account for more than 15.44% of all indexed literature. Regarding the authors, the highlight is Weisburd and Braga. As a result, the LDA grouped the latent words in the articles analyzed by themes studied and presented the list of articles by themes. The thematic map identifies the following themes as basic research subjects: community policing, problem-oriented policing, predictive policing, fear of crime and social control.

Practical implications

As the main implication between the combination of the bibliometric analysis method with the probabilistic topic modelling, is the emergence of a primordial step in the systematic literature review process, as this method allows to explore and group a large volume of data. Another practical implication that is intended is to provide the beginning researcher or any other reader with a panoramic view of the main authors who study the themes that impact police activity in any city in the world, which are the countries and reference centers of the study on the subject and, finally, the evolution of the main themes researched in the police area.

Originality/value

The value of these studies is summarized in the presentation of an overview on the theme in the last 50 years, offering the opportunity for other researchers to use this research as a starting point for other analyses.

SenU-PTM: a novel phrase-based topic model for short-text topic discovery by exploiting word embeddings

Purpose

Topic model has been widely applied to discover important information from a vast amount of unstructured data. Traditional long-text topic models such as Latent Dirichlet Allocation may suffer from the sparsity problem when dealing with short texts, which mostly come from the Web. These models also exist the readability problem when displaying the discovered topics. The purpose of this paper is to propose a novel model called the Sense Unit based Phrase Topic Model (SenU-PTM) for both the sparsity and readability problems.

Design/methodology/approach

SenU-PTM is a novel phrase-based short-text topic model under a two-phase framework. The first phase introduces a phrase-generation algorithm by exploiting word embeddings, which aims to generate phrases with the original corpus. The second phase introduces a new concept of sense unit, which consists of a set of semantically similar tokens for modeling topics with token vectors generated in the first phase. Finally, SenU-PTM infers topics based on the above two phases.

Findings

Experimental results on two real-world and publicly available datasets show the effectiveness of SenU-PTM from the perspectives of topical quality and document characterization. It reveals that modeling topics on sense units can solve the sparsity of short texts and improve the readability of topics at the same time.

Originality/value

The originality of SenU-PTM lies in the new procedure of modeling topics on the proposed sense units with word embeddings for short-text topic discovery.

Knowledge discovery in research on domestic violence: an overview of the last fifty years

Purpose

The database of the Web of Science (WoS) was searched for publications from January 1945–May 7, 2020 on the topic of domestic violence in titles, abstracts and keywords. The references were analyzed using the R bibliometrix package, and abstracts were analyzed using latent Dirichlet allocation (LDA) with collapsed Gibbs sampling to obtain topics related to domestic violence.

Design/methodology/approach

The aim of the study is to explore and provide an overview of research carried out on domestic violence, in its various aspects, over the past fifty years.

Findings

As a result of the research, the authors can assert that in the last fifty years, 32,298 authors have produced 19,495 documents on the theme of policing strategy and related subjects in 111 countries. Scientific production in this area grows at a rate of 12.81 per year. The United States of America is the leading country in publications with 48.14%, followed by the United Kingdom with 7.57% and Australia with 6.05%. Regarding universities, the highlight is the University of California with 664 publications, followed by the University of London with 515 and the University of North Carolina with 484. As for journals, the highlight is the Journal of Interpersonal Violence, Journal of Family Violence and Violence Against Women, which account for more than 14.32% of all indexed literature. Regarding the authors, the highlight is Campbell J.C and Feder G. Probabilistic topic modeling revealed that 18% of the topics concentrate 90% of all tokens. Topic 1 accounts for 27.9% of the sample and conducts research related to intimate partner violence.

Practical implications

As a practical implication of using the LDA in the bibliographic review, we infer that its capacity to explore large masses of data allows the researcher to explore an infinitely greater amount than the traditional methods of systematic literature review.

Originality/value

The value of these studies is summarized in the presentation of an overview on the theme in the last fifty years, offering the opportunity for other researchers to use this research as a starting point for other analyses.

Ranking policing strategies as a function of criminal complaints: application of the PROMETHEE II method in the Brazilian context

Purpose

The purpose of this study is modelling of a problem of policing strategy order using a multicriteria method.

Design/methodology/approach

For the construction of the impact matrix strategies under the reduction of crime rates, considering a portfolio of crimes, a questionnaire applied to specialists was used. In a second moment, defined the criteria and strategies to be ordered, the multicriteria PROMETHEE II method was used, which with the help of the Visual PROMETHEE software, emulated the systematised data in the impact matrix and produced the final ordering of the most efficient strategies, in the fight against crime, in the perception of decision makers.

Findings

As a result, this research revealed that radio patrol, when used in a non-randomised manner, is the most effective policing strategy in reducing the 18 criminal demands studied in the perception of decision makers after data emulation with the PROMETHEE II method.

Research limitations/implications

As research implications, it can be inferred that the use of multicriteria methods in the modelling of problems in public security area can contribute to the rationalisation of use of the available means in the fight against crime in large cities. This research showed that it is possible to use customised policing strategies to absolute reality.

Practical implications

The practical impact of this research lies in optimising the resources available to law enforcement agencies in the fight against crime in general.

Social implications

It can be inferred that by choosing appropriate strategies to combat local crime, there is a direct implication in optimising the resources that the government makes available to police agencies. This optimisation allows pressure reduction under the public budget for more features. The model for choosing more effective strategies contributes to local crimes decrease, increasing the sense of the population security.

Originality/value

The originality lies in filling a gap in the literature with the elaboration of the impact matrix of policing strategies in reducing criminal indices and in their associated use in ordering strategies through a multicriteria method. This study contributed to applied police intelligence.

Operational research applied in the field of public security

Purpose

Because that the crime in a wide way impacts the life of the people in the big metropolis, researchers have treated the question from several angles. The purpose of the paper, under the umbrella of operational research, is to develop a model of the ordering of police strategies, in the fight against crime in general, according to a certain criminal demand.

Design/methodology/approach

For the construction of the impact matrix of the strategies under the reduction of crime rates, considering a portfolio of crimes, a questionnaire applied to specialists was used. In a second moment, defined the criteria and strategies to be ordered, the multicriteria ELECTRE IV method was used, which with the help of the J-Electre software emulated the systematized data in the impact matrix and produced the final ordering of the most efficient strategies, in the fight against crime, in the perception of decision-makers.

Findings

As a result, the research revealed that policing strategies directed at solving specific crimes are the most effective in the perception of decision-makers after the emulation of data with the ELECTRE IV method.

Research limitations/implications

As research implications, it can be inferred that the use of multicriteria methods in the modeling of problems in the area of public security can contribute to rationalization of the use of the means available in the fight against crime in large cities. The research showed that it is possible to use customized policing strategies to a certain reality.

Practical implications

The method presented in this research is directly related to the major strategies: problem-oriented policing and hot spot policing. This method allows public safety managers to consider the possibility of combining different law enforcement strategies in each context. In this sense, the use of the multiple-criteria decision analysis (MCDA) (ELECTRE IV) method allows the evaluation of a large set of alternatives according to a set of established criteria, speeding up the process and reducing subjectivity, allowing the manager to analyze several scenarios with greater clarity and impartiality and choosing an alternative that best solves the proposed problem. The expected result is the rationalization of the available means applied in the search for the reduction of crime rates.

Social implications

The customization of policing strategies, according to criminal demands, implies the efficient way to reduce criminal charges. Reducing criminal rates enables the development of the local economy, tourism and the quality of life of people by exercising their freedom to the full.

Originality/value

The originality lies in filling a gap in the literature with the elaboration of the impact matrix of policing strategies in reducing criminal indices, and in their associated use in ordering strategies through a multicriteria method.

A model of policing strategy choice

Purpose

The purpose of this paper is to develop a method for the discovery of knowledge in emergency response databases based on police incident reports, generating information that identifies local criminal demands that allow the selection of the appropriate policing strategies portfolio to solve the problem.

Design/methodology/approach

The developed model uses a methodology for the discovery of knowledge involving text mining techniques using Latent Dirichlet Allocation (LDA) integrated with the ELECTRE I multicriteria method.

Findings

The developed method allowed the identification of the most common criminal demands that occurred from January 1 to December 31, 2016, in the policing areas studied. One of the crimes does not occur homogeneously in a particular locality. In this study, it was initially observed that 40 per cent of the crimes identified in the Integrated Public Safety Area 5, or AISP-5, (historical city center of RJ) had no correlation with AISP-19 (Copacabana - RJ), and 33 per cent of crimes crimes in AISP-19 were not identified in AISP-5. This finding guided the second part of the method that sought to identify which portfolio of policing strategies would be most appropriate for the identified demands. In this sense, using the ELECTRE I method, eight different scenarios were constructed where it can be identified that for each specific criminal demand set there is a set of policing strategies to be applied.

Research limitations/implications

The collected data represent the social dynamics of neighbourhoods in the central and southern zones of the city of Rio de Janeiro during the specific period from January 2013 to December 2016. This limitation implies that the results cannot be generalised to areas with different characteristics.

Practical implications

The developed methodology contributes in a complementary way to the identification of criminal practices and their characteristics based on reports of police occurrences stored in emergency response databases. The knowledge generated through the identification of criminal demands allows law enforcement decision makers to evaluate and choose among the available policing strategies, which best suit the reality they study, and produce the reduction of criminal indices.

Social implications

It is possible to infer that by choosing appropriate strategies to combat local crime, the proposed model will increase the population’s sense of safety through an effective reduction in crime.

Originality/value

The originality of the study lies in the integration of text mining techniques, LDA and the ELECTRE I method for detecting crime in a given location based on crime reports stored in emergency response databases, enabling identification and choice, from customized policing strategies to particular criminal demands.