Key Elements of Predictive Policing Include Crime Analysis, Crime Mapping and Geographies

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Key Elements of Predictive Policing Include Crime

Analysis, Crime Mapping and Geographies
Dr. John Motsamai Modise
South African Police Service
Abstract:- The goal of this article is to provide a officers at various levels (Perry et al., 2019: 3). The goal of
reference manual for those who are interested in writing PP is to "issue crime forecasts in the same way the Weather
on predictive policing, which will include evaluations of Service issues storm alerts," according to Zavrsnik (2017:
the most promising technical tools for producing 1), and so to disrupt the "production cycle" of crime. With
predictions as well as the most promising tactical automated justice, it will be possible to eliminate prejudices
strategies to act on such predictions. More generally, this and heuristics and limit essentially moral decisions to 'clean
research aims to place predictive policing in relation to and pure' mathematical reasoning (Zavrsnik, 2017: 1).
other contemporary, proactive policing measures:
Although predictive policing is merely a tool, it can be a The rise of predictive policing, which claims to be the
very helpful one. It is not a magic oracle. The second holy grail of policing by stopping crime before it occurs, is
section will go through how predictive policing is sweeping the country. According to Walter, Perry, McInnis,
conceptualized, as well as its potential and actual Carter, Price, Susan, Smith, John, and Hollywood (2013),
advantages and disadvantages. Review clarifies how predictive policing is the application of analytical
predictive policing is conceptualized, as well as its techniques, particularly quantitative techniques, to identify
potential, actual benefits, and disadvantages. likely targets for police intervention and prevent crime or
solve past crimes by making statistical predictions. It has
Predictive policing, also known as crime been common practice for many years to predict crime rates
forecasting, is a set of high technologies aiding the police using statistical and geospatial analysis.The algorithm
in solving past crimes and pre-emptively fighting and makes a forecast about the increased likelihood of a specific
preventing future ones. With the right deployment of crime at a specific location during a specific time period
such technologies, law enforcement agencies can combat using big data.
and control crime more efficiently with time and
resources better employed and allocated. The current The two main tenets of place-based predictive
practices of predictive policing include the integration of policing are that (1) mathematical forecasting
various technologies, ranging from predictive crime techniques can be used to predict future crime risk
maps and surveillance cameras to sophisticated in precisel y defined geographic areas, and (2) the
computer software and artificial intelligence. Predictive deployment of police resources t o those prediction
analytics help the police make predictions about where locations reduces the likelihood of crime occurring
and when future crime is most likely to happen and who (Bowers, Johnson, and Pease, 2004; Mohler et al.,
will be the perpetrator and who the potential victim. The 2011). Predictive policing can be defined as the application
underpinning logic behind such predictions is the of data analysis technologies by the police to generate and
predictability of criminal behaviour and crime patterns effectuate actionable forecasts of sources and spatiotemporal
based on criminological research and theories such as conditions of future crime. This definition implies that
rational choice and deterrence theories, routine activities predictive policing is a cross-cutting policing strategy, a
theory, and broken windows theory. multidimensional process encompassing not only the
generation of crime predictions by algorithmic-mediated
Keywords:- Predictive policing,Forecasting, Crime data analysis but also the gathering and preparation of input
mapping, Prediction, Pre-processing data and the “journey” of the prediction from the police
data,Algorithms,Policing technology, Social control, department to its implementation on the street (Perry et al.
Machine learning. 2013: 11-15; Bennett Moses and Chan 2018: 807).
I. INTRODUCTION In reality, the bulk of hotspot and place-based

predictive policing algorithms concentrate less on arrests
Police work is at a crossroads. There are more and and more on crimes like robberies, burglaries, andassaults
more cries for more equitable law enforcement. Predictive that are frequently reported to the police by members of the
policing systems have received a lot of attention as law public (Mohler et al. 2015 and Ferguson, 2017). The
enforcement officials adopt contemporary technology to objective is to dispatch police to locations where crimes
anticipate criminal conduct. Predictive policing has been have been reported by victims in order to stop further crimes
used in the majority of developed countries, while opinions from occurring there. The predictive narrative shifts the
on its efficacy are divided. Predictive policing began with police's attention away from what already occurred and
basic heuristics and algorithms, but as the technology toward what might occur, as well as toward the effective and
landscape has evolved, it has become more sophisticated. efficient use of available resources to combat crime (Beck
Results must be visible in order for these novel tactics to and McCue, 2009). Although the ideal situation for any
succeed, develop over time, and be applied by police rational police agency is to prevent criminal activity,
IJISRT23MAY2191 www.ijisrt.com 3312

ISSN No:-2456-2165
predictive policing also plays a crucial role in responding to cutting-edge computer applications," according to
criminal situations. Wikipedia. Crime prediction's main objective is to help the
police reduce crime, both strategically and tactically.
For instance, it might help the law enforcement agency Therefore, crime prediction in and of itself is insufficient
catch the offender in the act (Martens 2017). Although it has unless the results are applied to the decision-making
been suggested that predictive policing has been around for process, particularly when it comes to the deployment of
a while (Ferguson, 2020, Ferguson 2017, Bachner, 2013, personnel and resources.
Perry et al. 2013), it's possible that the importance of
cutting-edge technology in enhancing predictive policing is Predictive policing is described as "the application of
a more recent development. Prior to the shift toward using analytical techniques- particularly quantifiable techniques-
technology to anticipate crime, the police could only make to identify likely targets for police intervention and prevent
these predictions about crime based on their own expertise crime or solve past crimes by making statistical predictions"
and experience.However, in industrialized nations like the (Perry et al. 2013). The essence of predictive policing is best
United States of America, Geographic Information Systems summed up by Uchida, who is cited by Meijer and Wessels
(GIS) and Computerized Statistics (CompStat) were utilized (2019), who writes: "Predictive policing is a concept that is
to anticipate and address crime in the 1990s (Ferguson, et al. built on the premise that it is possible to predict when and
2017). The goal of predictive policing is proactive, as where crimes will occur again in the future by using
opposed to the old usage of GIS and CompStat, which was sophisticated computer analysis of information about
mostly reactive (Bachner, 2013). This means that crime previously committed crimes" (2019: 1033).This
should be avoided from happening in the first place. This information is utilized to anticipate potential future crimes
research, however, concentrates on modern advanced based on people, space-time, or both. In order to counter the
technology applications to anticipate crime, notwithstanding risks identified with appropriate police actions and to
the enormous contribution of these data-driven technologies prevent, decrease, dissuade, or at the very least capture
from the 1990s to deal with crime. criminals on the scene, it is necessary to forecast a future
crime as accurately as possible in time and location (Gluba
The transition to data-driven police operations, which & Pett, 2016: 2). Although the current emphasis is on
has been accepted recently and is expanding quickly, allocating police resources in response to threats that have
includes predictive policing (Brayne, 2017). Although there been identified, the ideal solution would be created to
isn't a single definition for predictive policing, the address the unique causes and circumstances of an issue and
fundamental traits can be summed up as follows. Predictive to encourage multi-agency action. Of course, it would be
policing is a police tactic in which a wide range of data necessary to consider the interests of particular groups as
about crimes is typically analyzed and visualized with the well as the effects on local communities and society at large
aid of software. A definition that reflects the core of when using data for multi-agency action.
technology in crime prediction is required because the main
idea of this paper—and maybe the future of predictive According to Chief Inspector of Constabulary Tom
policing—is based on cutting-edge technology. Perry et al.'s Windsor (as reported in Travis, 2013), the police have a
(2013) definition may be more precise and has been taken responsibility to stop crime from occuring and keep the
up by a number of authors.Meijer and Wessels provide a citizens of their community secure. Windsor argues that the
more recent definition, stating that "predictive policing is the lack of crime and disruption should be the key test for the
collection and analysis of data about prior crimes for police, not the obvious actions the police engage in in
identification and statistical prediction of individuals or dealing with it, in line with Sir Robert Peel, who established
geospatial areas in with an increased probability of criminal modern policing in 1829 (Emsley, 2011).So, while it may be
activity to help develop policing intervention and prevention a means to a goal, arresting criminals is not an end in and of
strategies and tactics" (2019: 1033). The concept is arguably itself. Since the public need not suffer injury first for the
a little convoluted, despite the fact that it seems to contain police to take action, this reinforces the idea of the social
the essential elements of predictive policing. The essential contract and increases police credibility. According to the
role that technology plays in crime prediction seems to be National Crime Prevention Council (1997), proactive
captured by contemporary definitions. policing is more closely associated with crime prevention
than reactive policing. Predictive policing, which is
Three key aspects can be derived from the above proactive and has gained popularity recently, may be the
definitions, and these are the importance of historic crime next development in policing strategy and the ultimate form
data, the essence of computer-based applications, and the of crime prevention.
anticipation of future crime. Another important observation
relates to the use of vast data from disparate sources, and we Traditional policing is not supplanted by predictive
want to emphasise, as shall be seen in this paper, that some policing. Instead, by processing pre-existing data more
of the data may not necessarily be relating to crime but other effectively and reliably, it offers the LEA and related
sociodemographic aspects. We give the following stakeholders (such as local governments, housing
condensed description of predictive policing after reviewing businesses, etc.) an extra tool. Ferguson (2017) asserts that
definitions from renowned academics and trying to discover there are three primary orientations for software-based
definitional convergence: "Predictive policing is a policing analytic approaches to predictive policing: (1) targeting
model that uses historical crime and socio-demographic data locations of property crime; (2) targeting locations of violent
from dispersed sources to predict future crimes using crime; and (3) targeting individuals engaged in criminal

ISSN No:-2456-2165
activity.The first is the most popular Predictive Policing II. SHORT HISTORY OF PREDICTIVE POLING
strategy, concentrating on carjacking, auto theft, and theft
from vehicles. They are currently regarded as the most Although techniques for anticipating crime have been
predictable crimes due to their frequency and the fact that around for a while, it has only been recently that
the majority of these crimes are reported to the police. contemporary technology has allowed these efforts to
Predictive policing may also be influenced by pertinent advance from straightforward heuristic techniques to
criminological ideas, such as van Eck's Crime Triangle (van complex mathematical algorithms. The LAPD and former
Dijk et al., 2015: 12). Furthermore, findings from social police Chief William J. Bratton are credited with developing
science research indicate that these crimes are motivated by the predictive policing concept. By 2008, Chief Bratton had
environmental vulnerabilities that can be found and made numerous public appearances to discuss the LAPD's
addressed with police actions, such as police presence in the accomplishments, including the department's recent
targeted neighborhoods (Ferguson, 2017: 1126–1127). adoption of predictive analytics to foresee gang violence and
assist real-time crime monitoring. According to Chief
Additionally, predictive policing developed into a Bratton (2015), this new strategy might build upon and
method for foretelling violent crimes in cities. Through improve upon current strategies like community-oriented
traditional hot spot policing, it is recognized that violent policing and intelligence-led policing.In 2008, Chief Bratton
crimes (including robberies, aggravated assaults, and collaborated closely with the acting directors of the National
shootings) frequently take place in particular places, such as Institute of Justice (NIJ), Kristina Rose and James H. Burch
nightlife areas where drugs and alcohol are used. II of the Bureau of Justice Assistance to examine the
Additionally, locations that have been claimed by criminal emerging idea of predictive policing and its implications for
organizations can reveal where gun violence is most likely law enforcement organizations. For this endeavor, the NIJ
to occur. Traditional hot spot policing characteristics were organized two consecutive predictive policing symposiums
modified and augmented with new data in order to forecast where leading researchers, practitioners, policymakers, and
violent crimes and establish a wider strategy (ibid., p. 1132- law enforcement officials gathered (Chief William Bratton,
1133).The perpetrators and victims came more and more 2015).
into focus in order to make even more accurate and detailed
predictions and to stop crime at its source. The method was Chief Bratton "served as the catalyst for bringing
modified for use in the field of predictive policing after predictive policing to the forefront," according to Kristina
being initially used to identify possible terrorist networks. Rose in her opening remarks at the first symposium, which
The fundamental idea behind forecasting criminals or was held in Los Angeles in November 2009. She also noted
victims entails a thorough examination of a person's social the interest in the sector as a whole in comprehending the
networks, previous criminal activity, or ties to known phrase "predictive policing" and the policy, technical, and
offenders using information from addresses, social media, operational ramifications of such approaches. She listed a
phone numbers, etc. A small percentage of the population number of law enforcement organizations from across the
has a higher probability of becoming a victim of crime or country, including the Baltimore State Police, Boston,
experiencing victimization repeatedly, according to Chicago, Los Angeles, D.C. Metropolitan, New York, and
victimization research findings. Shreveport police departments, who had responded to a
request for proposals from organizations interested in taking
Consequently, police tactics concentrate on a proactive part in a predictive policing demonstration initiative.At all
approach that directly addresses those who are most likely to levels of law enforcement, there was extensive discussion
commit crimes or become victims (Ferguson, 2017: 1138). about the Los Angeles symposium, and the subject of
The police's risk assessment of people and locations is predictive policing attracted a lot of attention on both
supported by this strategy. In order to incorporate them in traditional and social media. Quickly, consultants and
the study, it looks for prospective aspects that may not have private businesses started offering expert services and
been taken into account in conventional police work software that they believed would be valuable and
(Ferguson, 2017: 1125). According to Perry et al. (2013), appropriate for efforts at predictive policing.
effective measures to a particular risk need to be established
in order to have an influence on crime in the long run.A A second symposium was conducted in Providence,
small percentage of the population represented a large Rhode Island, in June 2010 with significant momentum.
percentage of the victims of crime, according to research Extended talks from the first symposium were presented
from the 1970s on the rates of re-victimization of people and during the occasion, and it was generally agreed that more
places (Sparks et al., 1977; Hindelang et al., 1978). The research into predictive policing was necessary. Major
same was true for specific locations. For a variety of points of emphasis included difficulties, achievements,
crimes, including burglary, domestic violence, bank robbery, restrictions, and scalability. Participants stressed how
and theft from motor vehicles, previous research have important it is to regionalize and share data as well as the
demonstrated the repeat victimization phenomena (Lamm- need of having excellent analytical skills (National Institute
Weisel, 2001; Braga & Weisburd, 2010; Johnson & Bowers, of Justice, 2012).Predictive policing, however, has a much
2004). longer history. US police departments started experimenting
with technical techniques for systematic data analysis in the
early 2000s. For instance, in order to analyze threats and
deploy tactical units based on risk, the Richmond Police
Department started using SPSS data mining programs in

ISSN No:-2456-2165
2003 (McCue and Parker, 2003; McCue, 2007). Later, hypotheses for this investigation into what we call a
comparable methods were adopted by other departments "blended theory":
(Robinson and Koepke, 2016).  Both criminals and victims have similar life patterns, and
overlaps in these patterns suggest a higher risk of crime.
The term "predictive policing" was first used explicitly  The where and when of certain patterns are influenced by
in 2008 (Perry et al., 2013: 4), and this framing was geographical and temporal characteristics.
steadfastly supported by the media attention that  As they follow these patterns, criminals choose whether to
accompanied the implementation of the software tool conduct crimes based on "rational" considerations that
PredPol by the police departments of Santa Cruz and Los include the environment, the suitability of the target, and
Angeles in 2011. PredPol is short for Predictive Policing. the likelihood of being apprehended.
The Los Angeles Police Department and University of
California researchers worked together to develop PredPol, Theoretically, we can uncover many of these trends
which was the result of their efforts to combine and characteristics using analytics, and we can then use
criminological theory and police data in order to create a tactical interventions to influence criminals' actions in order
practical analytical tool for police work (Mohler et al., to avoid crimes. Robberies, break-ins, and thefts are
2015). examples of "stranger offenses," which the mixed theory
best matches. Vice and relational violence, which both
The US National Institute of Justice (NIJ) and the include human ties that go beyond defined geographic limits
Bureau of Justice Assistance (BJA), which organized two and result in decisions that do not fit into conventional
seminal symposia on predictive policing in order to further "criminal rational choice" frameworks, are less appropriate
explore its potential, its organizational requirements, and its to this theory. Alternative explanations for vice and
potential effects on policing routines and practices, also domestic violence have, however, been put to the test. As a
played a significant role in the advancement of algorithmic result, tools and techniques for determining the dangers in
crime analysis methods (Pearsall, 2010). In order to ensure these areas have also been developed.
they received plenty of attention from the law enforcement
community, they consulted with William Bratton, a well- IV. PREDICTIVE POLICING AND OBJECTIVES
known figure in the American police scene and the former
commissioner of the police departments of Boston, New Results must be evident if we want to ensure that these
York, as well as former police chief of Los Angeles (Bureau novel tactics are adopted, develop through time, and are
of Justice Assistance, 2009; Perry et al., 2013: 4). used by police officers at various levels (Perry et al., 2019:
Additionally, to support basic and applied research on 3). According to Zavrnik (2017: 1), the goal of predictive
predictive policing, the NIJ provided funds to academics and policing (PP) is to "issue crime forecasts in the same way
police forces. When taken as a whole, these initiatives that the Weather Service issues storm alerts" and so to
provide a significant boost for the creation of crime disrupt the "production cycle" of crime. Automated justice
prediction software and its use in the US (Ferguson, 2017: offers the possibility of eliminating prejudices and heuristics
32). while limiting essentially moral judgements to "clean and
pure" mathematical reasoning (Zavrnik, 2017: 1). Predictive
III. THEORETICAL FRAMEWORK policing has the broadest breadth identified in the literature,
according to Pearsall (2010: 16).
The assumption that crime is predictable (in a
statistical sense) is well supported by a large body of data, In essence, predictive policing involves gathering data
mostly because criminals like to operate in their comfort from many sources, analyzing it, and applying the results to
zones. In other words, they frequently carry out the same or predict, stop, and handle crimes more skillfully in the future.
similar crimes that they have successfully accomplished in Utilizing algorithms to analyze vast amounts of data in order
the past. Although not always the case, this happens to predict and stop likely future crimes is known as
frequently enough for these approaches to function fairly predictive policing. Predictive policing seeks to reduce
well. Jeff Brantingham, an anthropologist at the University crime frequency and deploy police personnel with
of California, Los Angeles who assists in managing the consideration for available resources (Landeskriminalamt,
predictive policing project for the Los Angeles Police 2018: 10, Egbert 2017, 2018; Gluba 2016; Belina, 2016).
Department (LAPD), Despite the fact that Germany's overall crime rate is
dropping, this motive predominates in the federal states
The critics want you to assume that people are too (Knobloch, 2018: 10). Different prediction kinds are
complicated and unpredictable, and that this kind recognized in practice. On a superordinate level, predictions
of math is impossible. However, people are not between space and time or individuals can be distinguished
nearly as unpredictable as we imagine. Crime is and integrated in use (Egbert &Krasmann 2019: 12).
essentially just a physical process, and if you can
describe how criminals behave and interact with Using well-established criminological theories as a
their victims, you can learn a tremendous amount. foundation, predictive policing is a policing tactic that may
be incorporated into a larger pre-emptive policing approach
Major theories of criminal behavior, including routine (Ferguson, 2017; van Brakel, 2020; Egbert and Leese,
activity theory, rational choice theory, and crime pattern 2021). This tactic is based on the rationale that by
theory, are in agreement with Brantingham's statements examining known criminal behavior from the past, we may
(Clarke, and Newman, 2006). We combined these

ISSN No:-2456-2165
statistically forecast criminal behavior in the near future. Anticipation System in the Netherlands (Willems 2014;
This information can then be used to guide resource Drenth 2017), (Knobloch 2018; Seidensticker, Bode and
allocation and police methods (Brayne, Rosenblat, and Stoffel 2018; Egbert and Leese 2021) have all been tested by
Boyd, 2015). While this logic is at the heart of predictive European police agencies.
policing, not all predictive approaches are the same; as such,
this chapter will first explore three distinct types of The primary difference between these various tools is
predictive policing systems – hotspot policing, predictive in their ownership structures: some models were created
identification of individuals and predictive identification of internally by police, while others were created in
objects – before engaging with its critiques and exploring collaboration with universities, and yet others were
the relationship between society and the concept of risk as a purchased off the shelf from for-profit vendors. The way the
mechanism of governance. models are built is a further point of distinction. While most
studies calculate the spatio-temporal distribution of crime
In order to identify the areas where crime is most primarily using police data (the type, location, date, and time
likely to occur in the near future, hotspot policing looks for of the crime), others also take into account factors like
trends in the distribution of crime (Kaufmann, Egbert, and weather, holidays, events, and distance from highways
Leese, 2018). This type focuses mostly on foretelling high (Ferguson, 2017; Hardyns and Rummens, 2018), arguing
impact crimes like robberies, break-ins, and theft, and it can that certain seasonal factors and proximity to highways have
help determine how heavily police patrol certain regions. historically contributed to an increase in criminal activity in
The general consensus is that the presence of police in a certain areas.Given that police departments typically use
given area reduces opportunities for crime even in the location-based data, Hot-Spot methods are the most widely
absence of direct contact with potential offenders (Sherman used forecasting technology in the field of criminal
and Weisburd, 1995; Weisburd, 2008; Loughran et al. activities. According to Groff & La Vigne (2002: 34),
2011), according to experiments in hot spot policing (Braga "crime analysts prepare maps of crimes that have already
and Bond, 2008).This overall deterrent impact reportedly occurred and those maps are used to deploy officers and to
spreads into surrounding regions where the police were not identify areas in need of intervention" in this situation.
concentrating their efforts (Clarke and Weisburd, 1994) and Approaches for predicting criminals, the identities of
lasts for some time after the police have left (Koper, 1995; criminals, and possible victims of crime are more based on
Sherman and Weisburd, 1995). However, police patrols can person-related datasets.
also deter crime in other ways outside general deterrence.
By physically immobilizing potential offenders, direct If deployed and taught properly, the usage of
interference through stops, searches, detentions without intelligent prediction tools can offer a number of
arrest, and arrests may deter crime (Weisburd and Eck, advantages. To help investigators identify, foresee, and
2004). resolve criminal investigations, the analytical function
creates a variety of intelligent products. Thus, prosecution is
If prolific offenders are being arrested, this use of based on information that has been gathered and presented
selective incapacitation may have an immediate impact on in carefully designed tables, charts, maps, or other
crime (Wyant et al. 2012). If those repeat offenders are later graphics.Aiming to support both the decision-making of top
expelled from the community, incapacitation might have executives or the agency's goal (Ioimo, 2018: 6). Supporting
longer-term consequences. There is a lot of evidence to the adjudication of trials. Tactical and strategic suggestions
support the idea that bias, both verbal and unconscious, can can help law enforcement officers. These reports can include
significantly affect who is stopped, searched, and crime hotspots, crime bulletins, and crime summaries, as
arrested.Predictive policing raises legitimate concerns that it well as crime trends, potential threats, vulnerability, and risk
might increase these prejudices and strengthen any assessment analyses (Iomio, et al. 2007: 7).
propensity for police to target minorities and their
neighborhoods (Ferguson Citation in press). Such worries By facilitating quick access to information and
arise even if the forecasting techniques used to power assisting in multijurisdictional cases, computerized
predictive policing do not include information that would databases serve as a foundation for PP, organize
explicitly be biased. Any crime reduction advantages would information, and establish meaningful ties with other law
need to be considered in terms of their discriminatory costs enforcement staffs. The outcomes of predictive policing are
if predictive policing unintentionally worsens bias. In the in compliance with local, state, tribal, and federal laws and
worst-case scenario, proven benefits might only result from regulations since the technologies are tailored to the relevant
prejudice brought on by projections. legal circumstances (Ioimo, 2018: 7). Additionally, existing
resources can be used more effectively and resources can be
In other words, projections without this bias wouldn't conserved over time by utilizing underpinning software.
result in any reductions in crime. In reality, most hotspot and This could also suggest that police officers are employed
place-based predictive policing algorithms (Mohler et al. properly rather than necessarily making staff savings.Place-
2015; Ferguson, 2015) concentrate less emphasis on arrests based predictive policing, the most often utilized method,
and more emphasis on crimes like robberies, burglaries, and frequently examines pre-existing crime data to identify
assaults that are frequently reported to the police by regions and times with a high possibility of crime. Predictive
members of the public. The objective is to dispatch police to policing, according to some, might help police departments
locations where crimes have been reported by victims in save money by making crime-reduction programs more
order to stop further crimes from occurring there. The Crime effective. "Predictive policing is a more rapid means of

ISSN No:-2456-2165
combating crime."The public values crime prevention more police activity that uses well-known, crime-relevant
than responding to crimes and apprehending offenders. variables to identify hotspots as they emerge.
Although it is occasionally erroneously claimed that
forecasts have little to do with crime prevention and are only Thus, the collecting of data is based on either person-
useful to "catch" criminals, (Schwabe, & Schurink, 2000). or space-related data. Methods for forecasting crimes are
included in location- and time-related forecasts; these
Systems for predictive policing, for instance, are techniques are used to identify locations and periods that are
capable of quickly processing massive amounts of data. more likely to experience a certain criminal behavior (Perry
Police officers now have more time to devote to other tasks et al., 2013, p. xiv).Aiming to pinpoint people who will
like street patrol. The adoption of modern forecasting commit a crime soon, approaches for predicting offenders
technology can streamline investigative processes in this go into detail. Profiles that precisely match likely offenders
way, and in the best case scenario, crime rates can be with certain historical crimes are the subject of research into
decreased through effective preventative measures.As a identifying the perpetrators. It is possible to identify groups
result, the following areas can benefit: police personnel or individuals who are more likely to become victims of
management, such as professional deployment and offenders using tools that predict probable crime victims
recruitment; police budget management, such as measuring (Perry et al. 2013, p. xiv).Despite the fact that forecasting
the costs of overtime and other expenditures; offender capabilities can be divided into many groups, the techniques
monitoring; city or neighborhood planning, such as design for subsequent application and execution are comparable.
of spaces; police security resource allocation; or According to Perry et al. (2013), the PP process can be
infrastructure protection (Uchida, 2010). For instance, the described as a traditional four-step comprehensive business
LKA of the federal state of North Rhine-Westphalia cycle. as illustrated in picture 3. Data fusion is necessary for
established the following goals for German police the first two processes, which deal with gathering and
organizations during the PP project's implementation phase analyzing crime, incident, and offender data.
dubbed SKALA: Its goal is to conduct targeted, strategic
Fig. 1: Prediction-led process (Perry et al., 2013)
Four sections make up the cycle of gathering data, the identities of criminals, or the identities of crime victims.
forecasting outcomes, and putting those outcomes into The strategic subject list (SSL) in Chicago, where a risk
practice (see Fig. 1). The first step is the thorough gathering score was assigned to people who had previously been
of quantitative data, which is necessary for precise forecasts; arrested, predicts a heightened probability of becoming
the second step is the actual analysis of the data; the third engaged in future gun violence either as a perpetrator or a
step is the prediction of crime; and the fourth step is victim, is the most notable example of "person-based
particular police actions that aim to prevent crime by predictive targeting". The SSL had low predictive accuracy
deterring offenders or arresting them. in the aforementioned "prediction-led policing business
process," in addition to lacking instructions on how to
Accordingly, "location-based or geospatial predictive integrate predictions into police operations (Saunders, et al.
policing" and "individual-based predictive policing" are the 2016).This finding emphasizes the need to consider the
two subcategories that are typically used to separate out entire "prediction-led policing business process" when
predictive policing (Sommerer, 2017). According to examining and evaluating predictive policing—whether it be
Ferguson (2017), "person-based predictive targeting" and "place-based predictive policing" or "person-based
"place-based predictive policing" are subcategories. predictive targeting"—because each strategy can produce its
According to Perry et al. (2013), the latter category can be own distinct if faculties throughout the process.
further separated into techniques for predicting criminals,

ISSN No:-2456-2165
Only a few attempts have been made to evaluate phase involves analysing the data based on the selected
implemented predictive policing strategies from an probability model. For example, the forecast calculation
independent scientific perspective (for example, Saunders et points out, which offence with an increased risk will take
al. 2016; Mohler et al. 2015). This is despite the fact that place in which area. Within the fourth step results of the
many police departments in various countries have been prognosis are presented graphically for police officers and
testing and using varying predictive policing strategies and investigators. This can be done with proper dashboards on
that police statements and media reports frequently sound smartphones or tablets in order to utilize them ad-hoc.
promising. Researchers from Germany (Gluba 2016; Belina
2016; Egbert 2017) and the US have also critiqued this The data can serve as a foundation for decision
(Perry et al. 2013; Ferguson 2017; Uchida 2014). management depending on the present situation. Police
officers can then put precise preventative measures into
Thethird step entails actions taken by police action. The last stage summarizes the applied modeling
institutions to stop the anticipated criminal activity or even performance measurement, retains lessons learned, and
to solve past crimes. The complexity of potential confirms the veracity of intermediate results (Bode et al.,
interventions is categorized into generic, crime-specific, and 2017: 3). In contrast to Perry et al.'s Prediction-Led Policing
problem-specific components. Complex treatments, approach, Bode et al.'s adaptation makes use of the
according to Perry et al. (2013, p. xviii), involve more feedback-culture and conducts continual evaluations at the
resources, such as personnel, but produce better, more goal- end of each phase.However, if the goal is to catch a criminal
oriented results. Managers should not only describe the in the act, for example, some forms of reactive policing may
crucial component of preventative analysis but also provide be better handled by predictive policing. Predictive policing,
information that satisfies the requirement for situational sometimes known as "crime forecasting," is the process of
awareness among officers and personnel in order to carry identifying or anticipating prospective criminal activity
out missions successfully (Perry et al., 2013, p. xviii).The using statistical and analytical approaches.
cycle's fourth stage can be finished by building on this. Each
intervention results in a criminal response that, in the ideal The Santa Cruz police were the first law enforcement
scenario, reduces the danger or stops the crime. Here, a organization to employ predictive policing. Numerous
short-term feedback and assessment is taken into account by police departments all across the world have implemented
ensuring that the interventions are being implemented predictive policing in recent years, which uses statistical
properly and that there are no obvious concerns. data to inform decision-making (Meijer, 2019). Predictive
Reprocessing the newly obtained data following each analytics and powerful computers are used to forecast
operation is required for Predictive Policing to be effective crimes including auto thefts, property crimes, and burglaries.
over the long run, which in turn results in shifting Thus, predictive policing poses some significant issues
environmental conditions. regarding the nature of prediction in a time where data
gathering and analysis are pervasive (Ferguson, 2017).
Belina, (2016: 1-2) contend that this illustration does
not fit the methodological standards as it is implemented in As a whole, predictive policing appears to be the next
Germany, despite the fact that Perry et al.'s Prediction-Led step in the evolution of policing strategies rather than
Policing Business Process is meant to represent a replacing officers or substituting existing policing strategies
comprehensive approach. The process is shown in the or models like "compstat," "evidence based policing,"
accompanying image, which was altered by scientists "problem oriented policing," "community oriented policing,"
(Seidensticker, 2017: 96) and shows it from the viewpoint of or "intelligence led policing" (Ferguson, 2013; Pearsall,
the police (Belina, (2016:.2).The resources will determine 2010). One can argue that Predictive Policing is only a
whether or not the use of predictive policing methods is component of a comprehensive strategy to combat crime in
advantageous in the specific setting. As a result, it is light of the fact that it does not always address the root
essential to begin with a delinquency analysis and determine causes of criminal behaviour.
whether applications are appropriate (Seidensticker, 2017:
296). If the usage of the Predictive Policing tool proved According to Smith (2014), "Predictive policing"
worthwhile in this situation, the executors provide includes the following features:
comments at the conclusion. Consequently, variations across  Predictive policing has been practiced by crime analysts,
federal states are always possible. The first step starts the police personnel, and to some extent the general public for
gathering and processing of datasets as well as the many years—since the very beginning of policing. People
examination and selection of data records. have generally always been able to make rational
"predictions" about the likelihood of the next crime by
Consolidation of space and time becomes essential. using historical data about past crimes, the location, the
Police-related data that has already been recorded can be time, and the criminal's behaviour. Even if modern
integrated with non-police data, including changes in the technology can improve our ability to anticipate crime, in
weather or temperature. It is crucial to geographically earlier times, the police relied on their own expertise and
reference data for this use in order to ensure a consistent, experience to make these predictions.
machine-process able dataset that serves as the foundation  Predictive policing has been practiced for many years in
for subsequent analysis. In the second phase, a specific some labour-intensive forms (Perry, McInnes, Price,
statistical model is built using the available data, such as a Smith, & Hollywood, 2013). Despite the fact that some
regression or decision tree (Box, et al. 2015: 305).The third people think it's a recent development. However, it is now

ISSN No:-2456-2165
possible to generate analytical data in real-time because to Additionally, Predictive Policing may involve
technology advancements (Smith, & Hollywood, 2013). allocating resources in response to anticipated increases in
Predictive policing is centered on data analysis. The crime. It is possible to identify specific crime patterns or
literature frequently claims that predictive policing predict the mass release of criminals from jail (Smith, 2014).
involves using algorithms that use sophisticated data However, in actuality, the officer may have altered the
analysis, geospatial information, complex mathematics, surroundings in response to the predicted counsel. Crime is
and, in some cases, anthropological and criminological deterred by the consequences of being in a potentially
behaviour research (Ferguson, 2013; Pearsall, 2010; Perry criminal environment, which may even result in fewer
et al., 2013).The objectives of predictive policing are to arrests (Smith, 2014). As the potential offenders are not
predict, prevent, and minimize crime, to effectively apprehended but rather the potential crime is prevented, this
address crime, and to define realistic goals for may give the police officers an instinctive sensation of
interventions (Perry et al. 2013; Pearsall, 2010. Some failure.
people are under the misconception that predictive
policing will result in more arrests rather than fewer. Boba, (2009), had actually claimed that "do[ing] more
 Predictive policing systems not only employ data from with less" would be the key to innovative approaches to
various sources but also information from the police crime analysis and prevention. In this context, technological
regarding crime. The choice of data can be problematic in innovation was widely seen as a potential means of
and of itself, and ethical concerns regarding the use of enhancing police work through more targeted and informed
predictive policing in the United States are becoming ways of acting (Beck and McCue, 2009; Bratton et al., 2009;
more and more prevalent (Pearsall, 2010). Saunders et al., 2016).
 "Predictive policing is the collection and analysis of data
Thus, from its inception, the foundation of predictive
about prior crimes for identification and statistical
policing was largely based on commercial logics. Predictive
prediction of people or geographic areas with a higher
policing has been characterized as an internal management
probability of criminal activity to help develop policing
tool that harmonizes police work with contemporary
intervention and prevention strategies and tactics." Albert
business processes in order to boost efficiency and
Meijer and Martijn Wessels, 2019: 3.
effectiveness (Saunders et al., 2016), which is one
 Predictive policing tactics are enticing LEAs, their interpretation of the practice that some have advanced in
partners, and other stakeholders in crime prevention to light of these motives. Others have highlighted the purported
examine the gathering and potential use of various data impartiality of algorithms as a motivating factor (Shapiro,
sources more critically. Predictive policing has created 2019).As stated by Ferguson (2017: 21ff), the adoption of
ethical questions in the US, though, and its use in predictive policing in the US was in reality influenced by a
European contexts has raised some worry. number of factors, not the least of which was the rising
Then, how does predictive policing differ from hostility between law enforcement and racial and ethnic
conventional approaches and police tactics? Some hints can minorities, particularly in the wake of the Black Lives
be found by examining the various definitions that Matter movement. In this light, predictive policing was
practitioners and academics have come up with for viewed as a technique to eliminate human prejudice and
predictive policing. Predictive policing, according to allow a machine to pick who to regulate and where to patrol.
Bratton, et al. (2009: 3), is "forward-thinking crime Of course, there is no such thing as an impartial algorithm,
prevention" that "connects technology, management and numerous studies over the past few years have shown
practices, real-time data analysis, problem solving, and how bias creeps into algorithmic decision-support systems
information-led policing to lead to results" such as crime in policing and criminal justice in a variety of ways (Angwin
reduction, effective agencies, and contemporary and et al., 2016; Lum and Isaac, 2016; Richardson et al., 2019).
innovative policing. The focus of the literature review will be on two ends
Predictive policing is defined similarly by Uchida of a spectrum that can be characterized in accordance with
(2009: 1) as a "multidisciplinary, law enforcement-based the proposed definition of predictive policing. Statistical
strategy that brings together advanced technologies, analytics, which uses complex arithmetic and algorithms to
criminological theory, predictive analysis, and tactical forecast future crime, might be seen as being at one end of
operations that ultimately lead to results and outcomes - the spectrum. Intuition, which is the other extreme of the
crime reduction, management efficiency, and safer spectrum, is unconscious human analysis based on
communities." experience.The technical, economic, and political pathways
that predictive policing adopts are numerous. Algorithmic
"Massive crime reductions following increased arrests crime analysis techniques are frequently advocated as a
will necessarily result from accurate predictions (within sophisticated solution to organizational problems and
predictive policing)," This clarifies the concept of predictive outside influences. Therefore, it is not unexpected that over
policing since it involves more than simply analysis and also the past ten years, predictive policing has extended swiftly
involves police action. The only output of using crime data to numerous local and national contexts all over the world.
analysis to make forecasts is, in fact, just a prediction. Police
must take action, with all the complications that entails, in
order to minimize crime (Perry, McInnes, Price, Smith, &
Hollywood, 2013).

ISSN No:-2456-2165
In conclusion, approaches that are more reliant on again within seven days increases initially but then
person-related datasets include methods for forecasting decreases.
criminals, the identities of criminals, and possible victims of
crime. Methods for predicting offenders specifically try to The analysis of geography and time serves as the
identify those who will likely commit a crime in the near foundation for resource deployments. A Decision Support
future. On the other hand, identifying the perpetrators System (DSS) was created by Camacho-Collados and
focuses on profiles that precisely match likely offenders Liberatore (2015) in conjunction with the Spanish National
with certain past crimes. Police Corps to effectively disperse police officers across a
region. They attempted to provide a solution to the Police
Groups or individuals that are more likely to become Districting Problem (PDP), which is the issue of how police
victims of an offender can be identified using tools that officers can best identify the patrol sectors where there is the
predict probable crime victims (Perry et al. 2013, p. xiv). greatest likelihood of criminal activity.
Despite the fact that forecasting capabilities can be divided
into many groups, the techniques for subsequent application The authors' DSS-model can be used to more
and execution are comparable. According to Perry et al. effectively assign police officers and choose the best patrol
(2013), the PP process can be described as a traditional four- routes. The authors' system was put to the test as part of
step. Data fusion is necessary for the first two processes, their research, and they came to the conclusion that it could
which deal with gathering and analyzing crime, incident, deploy police troops more effectively across the city.
and offender data. The data will be analyzed in accordance Although this approach appears to be very promising, it still
with the various police activities and divisions to make needs to be used in the real world before its true worth can
precise predictions. be assessed.
V. POSSIBLE ADVANTAGES OF PREDICTIVE Predictive police tactics are said to assist in identifying
POLICING people who may be involved in criminal activity, either as a
victim or an offender, in a second specific claim. According
Law enforcement organizations use these methods to to Perry (2013), predictive algorithms can be used to
more effectively and efficiently use their resources, which is identify members of criminal organizations that have a
how predictive policing is envisaged with broad potential higher than average likelihood of engaging in violent
benefits built in. Andrejevic, 2017) claim that predictive conflict (such as gang shootings). Inductive profiling is
policing can enable police forces to intervene by identifying another tool that can be used to identify people who may
patterns in vast data sets. We reviewed the literature and later commit crimes (Van Brakel & De Hert, 2011). With
identified the specific claims. the use of these methods, those who exhibit characteristics
that are associated with a higher likelihood of engaging in
The more accurate allocation of resources in both criminal behavior can already be watched or targeted (Perry,
space and time is the first distinct advantage of predictive 2013). This profile goes beyond only looking at an
policing. To identify high-risk areas, predictive policing individual's demographics; it can also include their (social)
methods that employ both previous crime data and a wider behavioural habits (Van Brakel & De Hert, 2011).
range of data are used. For example, risk terrain analysis and
advanced hot spot identification techniques are used to Kump et al. (2016: 156) used his research to show that
identify the areas where criminal activity is most likely to sex offenses are most commonly clustered in the offenders'
take place. Both criminal data and data from data mining are activity zones (i.e., places where people frequently go).
crucial for this geospatial analysis: they are data that don't Based on a more in-depth investigation, Kump et al. showed
immediately matter but could help with crime prevention through social network analysis that people are more likely
and prediction (Andrejevic, 2017). to commit crimes for a length of time (about 25 weeks) if
they are socially related to an offender to a certain extent.
These numerous data sources can also be used to Similar conclusions on social media and criminal activity
identify the times of day when criminal behaviour is most are reached by Williams, Burnap, and Sloan (2016: 337),
likely to occur through the use of spatiotemporal analysis. who discover a correlation "between aggregated open-
These algorithms are intended to forecast the peak period of source communications data and aggregated police recorded
criminal activity in a specific area. However, the research crime data in London."
also shows more specific patterns. Near-repeat crimes are
the idea that following crimes are more likely to happen VI. PITFALLS/ DRAWBACKS OF PREDICTIVE
close to the time and place of current crimes. POLICING
Dario, Morrow, Wooditch, and Vickovic (2015) Predictive policing can bias police personnel by
looked into the connection between crime and favourable spotlighting high and low-risk neighborhoods, placing them
surfing conditions in California (i.e., weather conditions that on unnecessarily high alert or giving them a false sense of
entice locals, visitors, and surfers to surf areas). The Ventura security. Predictive policing has an effect on neighborhood
Police Department's historical criminal data was used. They residents in addition to criminals and law enforcement
determine that the weather does, in fact, enhance crime in personnel. While many academics and police departments
these areas, but only from 2:30 to 5:29 pm (Dario et al., are persuaded that predictive policing has a bright future,
2015: 271). Similar conclusions were reached by Haberman some academics have expressed some reservations about the
(2017), who noted that the risk of armed robberies occurring use of data mining and algorithms to anticipate criminal

ISSN No:-2456-2165
conduct. These issues will be examined, and it will be et al., 2015). In their study, Schlehahn et al. (2015) offer a
established whether or not they are grounded on fact or fictitious illustration of how the acts of law enforcement
merely on speculative assumptions. We only mention this agencies might affect the resocialization of ex-
distinction if it is particularly highlighted because many of offenders.They present a convincing case for how the
these possible downsides apply to both the spatial-temporal stigmatization of some categories of criminals might cause
forecasts and the profile. aversion and, ultimately, relapse in criminal behavior as
their reintegration into society is slowed down by these
Due to the predictive policing models' lack of prediction algorithms and how officers handle them.
openness, it has been noted in the literature that law Because it encourages criminal activity, individual profiling
enforcement agencies are unable to completely understand may ultimately have a self-fulfilling outcome.
the algorithms (Datta, Sen, & Zick, 2016; Schlehahn et al.,
2015). Riskier is not the same as risky (Saunders et al., Predictive policing practices may sometimes have
2016), hence it may be difficult to assess how risky unforeseen repercussions when used. A startling result was
geographical areas or specific individuals are if the models reached by Edwards and Urquhart (2016) after comparing
are not understood by law enforcement organizations. The two data-driven programs, one of which dealt with
efficiency of law enforcement efforts may be compromised predictive policing. He examined the Kansas City No
if they are unaware of the elements that raise the likelihood Violence Alliance (KCNOVA), which employs network
of crime (Perry, 2013).In order to establish appropriate analysis software to pinpoint those who are most likely to
strategies, law enforcement organizations must also ensure engage in criminal activity, and a living lab in Kansas's
that the data is properly interpreted and that appropriate downtown that aims to raise living standards and encourage
inferences are made from it (Townsley, et al. 2023). capital investment in this area of the city.According to
Brannon (2017), the use of predictive policing in one
Predictive models are typically data driven rather than geographic area of the city has an effect on both this space
theory driven, which mayhave significant effects on how and its occupants: when one area is keeping an eye on
these models are applied. The use of big data and data-based criminal activity while the other is thriving due to the
methodologies could result in an overemphasis on encouragement of capital investments, this causes spatial
correlations rather than causality (Andrejevic, 2017). This inequality across racial and social classes.
might be a concern because forecasts based on algorithms
are ambiguous and challenging to understand (Chan & More basic concerns about privacy and ethics are
Bennett Moses, 2016). The models will be out of date and raised after the practical challenges that come with the use
give a distorted view of reality if they are not analyzed and of predictive algorithms. Edwards and Urquhart (2016)
appraised using practical insights such as the tacit examine whether it is appropriate for law enforcement
knowledge of police officers (Perry, 2013). authorities to use social media and open source data, as well
as how well protected citizens' digital identities are. The
As previously mentioned, the Saunders et al. (2016) writers dispute the degree to which citizens' digital
study also suggests that predictive algorithms might not be footprintssuch as what they post on social media and data
self-explanatory. Although the interaction with potential that can be gathered about our transit movements—are
offenders increased, one of the reasons they did not find private and whether it can be utilized indefinitely.
substantial results may be because the models do not offer
enough guidance on how to deal with these offenders or how The difficulties between social profiling of individuals
the models should be applied. This supports the idea that the and legal protectionswhich are frequently relaxed to address
predictive models can never be utilized independently these tensionsare discussed by Lammerant, & de Hert,
without additional guidance for police officers regarding (2016).Even while the legal limits of predictive profiling are
how to behave in public, which reduces their efficacy. fairly obvious thanks to privacy law, there are still a few
isolated instances that make it difficult to draw clear lines
Accountability issues could develop if prediction around who is qualified. Costanzo, D'Onofrio, and Friedl
models are not transparent or understood. Bennett Moses (2015) highlight this conclusion by arguing that regulation is
and Chan (2016) bring up the potential repercussion that law crucial to maintaining confidence between citizens and
enforcement may not be able to properly comprehend and governments since there should be a balance between the
interpret the software's results and may not view them as use of big data and citizens' privacy. Citizens may grow
sufficient information for making decisions. This can result deeply suspicious of governments if there are no limitations
in an accountability gap if police officers are unable to because they are unsure of whether and how much they are
recognize flaws in the models because they do not being watched (Schlehahn et al., 2015).
comprehend them. In other words, when complete
dependence is placed on predictive algorithms, it becomes Predictive policing can bias police personnel by
unclear who is in charge of making decisions. spotlighting high- and low-risk neighborhoods, placing them
on unnecessarily high alert or giving them a false sense of
Because of the lack of transparency, using a predictive security. The following predictive policing pitfalls affect
policing model for profiling may stigmatize people and neighborhood people other than criminals and law
groups, leading to automated forms of prejudice. When enforcement personnel:
predictive models are applied incorrectly, law enforcement
officials may overlook and underestimate the effect since it
may result in the stigmatization of some people (Schlehahn

ISSN No:-2456-2165
 The researchers asked respondents in a study that was with police patrols can have a direct detrimental effect
published in the British Psychological Society whether while not necessarily reducing crime. Overall, it is
they felt more or less comfortable among police officers. unethical to utilize predictive policing since it makes
They came to the conclusion that "the presence of a police residents feel insecure in their areas while police are there.
officer could serve as a warning signal, drawing attention  This might also go the other way; in a low-risk area, an
to possible danger in the area." officer might mistake suspicious activity for nothing.
 As a result of their use of skewed data, the existing Predictive policing can bias police personnel by
applications of predictive policing contradict the moral spotlighting high- and low-risk neighborhoods, placing
principles of justice and fairness. them on unnecessarily high alert or giving them a false
 Regardless of the location of the police, if there are more sense of security.
police on the streets, criminal activity will be reduced. As  Examining the effects of this technology is crucial to
a result, the success of predictive policing cannot be comprehending predictive policing in the context of this
directly linked to its success. The opposite of what is ethical framework. Predictive policing, from the
theoretically intended, though, may also happen as law standpoint of the police, helps to pinpoint potential crime
enforcement activity rises in a neighborhood. Ironically, hotspots. However, even if initiatives like LASER may
the increased police presence might make people feel produce promising outcomes, predictive policing may not
unsafe by instilling fear. necessarily be to blame for any particular drop in crime
 This ethical framework is one useful technique to assess rates.
the efficacy of predictive policing and determine whether
the strategy is morally acceptable. Based on a big database Predictive policing has an effect on neighborhood
of past crime data, predictive policing predicts where residents in addition to criminals and law enforcement
crime will most likely occur. The previous arrests must be personnel. Some may believe it is true that more crimes are
impartial in order to produce impartial forecasts because being committed voluntarily by residents of high-risk areas,
the software relies on outdated data. When crucial and that this justifies the increased police presence in certain
information and context are missing, crime data may be areas. However, this viewpoint makes generalizations about
skewed in one way or another. an entire neighborhood or community based on the few
 Data on past crimes can also reveal racial bias. For the criminals that actually exist.
same crime, persons of colour are arrested much more VII. KEY FINDINGS OF THIS ARTICLE
frequently than white people. Compared to white
individuals, black people were five times more likely to be  Predictive policing is the application of analytical
arrested by the police. In communities where there are techniques to identify promising targets for police
more arrests of persons of colour, racially skewed arrest intervention, with the goal of reducing crime risk or
data produces racially biased forecasts. For instance, even solving past crimes.
though drug usage rates "are essentially the same across  Predictive policing techniques can be used to identify
Oakland neighborhoods," a simulation research on places and times with the highest risk of crime, people at
predictive policing in Oakland, California, included a risk of being offenders or victims, and people who most
prediction algorithm that advised more police should be likely committed a past crime.
sent to neighborhoods with a high concentration of Black  To be effective, predictive policing must include
and Latinx residents. interventions based on analytical findings. Successful
 Predictive policing encourages overpolicing in interventions typically have top-level support, sufficient
communities of colour, which serves to legitimize the resources, automated systems to provide needed
fundamentally unjust structural racism that afflicts this information, and assigned personnel with both the
nation. freedom to resolve crime problems and accountability for
 As a result of predictive policing, more people are doing so.
arrested, which exposes them to "countless fines, fees, and  Many agencies may find simple heuristics sufficient for
other costs" that they frequently cannot afford. their predictive policing needs, though larger agencies that
 In addition, people with criminal records are unable to collect large amounts of data may benefit from more
vote, obtain a driver's license, or look for employment. complex models.
 In general, the justice system unfairly prevents persons of
colour and those living in poverty from re-entering society VIII. RECOMMENDATIONS
after committing crimes. Law enforcement can
increasingly target those who are already having difficulty The knowledge gained from these other domains can
thanks to predictive policing. be used to highlight the numerous known traps that might
 Victims of predictive policing are chosen based on bias, mislead practitioners and offer suggestions to improve the
and those who employ the practice may develop deadly efficacy of these techniques:
bias. If an officer is aware that they are in a high-risk area,  Concentrating on Prediction Accuracy Rather Than
they may interpret an activity that is not typically Tactical Utility: Using research from RAND into
suspicious as threatening. counterinsurgency operations in Iraq as an example, a
 Predictive policing has a negative effect on this safety computer model predicted that an IED event will take
issue. Even if there are more police patrols, more place somewhere in the city of Mosul within the following
individuals feel frightened as a result. Inundating a town 48 hours. While true, it information was rarely tactically

ISSN No:-2456-2165
useful. Practitioners should generally concentrate on  Neglecting Civil and Privacy Rights: Taking action in hot
generating conclusions that have tactical utility while locations is easier than worrying about civil and privacy
conducting tactical analysis. This means that the analysis's rights when using comparable methods to locate "hot
scale should correspond to the range of potential people." What should be done when a parolee is predicted
responses. A beat officer, for instance, is probably able to to have a high probability of reoffending even if the
handle a few hotspots the size of a city block but would forecast is still far from certain despite being much better
not find it feasible to concentrate on a strip several miles than chance? The typical response up to this point has
long. been that law enforcement is mostly free to conduct
 Relying on Bad Data: Systematic mistakes in the data will legitimate investigations and take preventative measures
result in systematic mistakes in the analysis. It will look against high-risk individuals because the majority of them
that there is no crime in particular places if data are are already under some type of correctional supervision
excluded. Since department analysts can identify issues in (or are at the very least convicted offenders. Again, it's
the output that would cause these kinds of systemic errors, unclear exactly what being "reasonable" means and under
it is crucial for police departments to comprehend the what circumstances. We think that over the next few
ground truth when doing these studies. Positional years, there has to be a lot more study and development
accuracy of geocoded crime events is a crucial element in done in this area.
predictive crime mapping, according to a research of more
than 400,000 criminal incidences from six sizable law IX. CONCLUSION
enforcement jurisdictions in the United States.
One of many policing tactics will continue to be
 Misunderstanding the Prediction's Causing Factors: Using
predictive policing. It is necessary to temper the lofty
common sense to verify the elements included in the
ambitions of this relatively new subject with pragmatic
model before applying techniques like regression or any
realism. To maximize the return on the necessary
of the data mining variations will assist avoid erroneous
technological investment, predictive policing must also be
correlations. Additionally, it's critical to be curious about
properly examined, monitored, and adjusted. Predictive
the factors that influence forecasts in order to uncover
policing models must take into consideration a variety of
deeper connections. For instance, comparing crime
factors because crime is a complicated phenomenon that is
statistics to police officer locations would probably reveal
influenced by many different things. This might affect the
a strong association between where crimes occur and
accuracy of forecasts. Police data (and other data sources)
where police have been stationed.This does not imply that
predictions are simply risk probabilities and do not offer a
we can determine where crimes will occur by looking at
clear window into the future.
the position of police, as officers are typically called to a
scene after a crime has already occurred. This association Predictive policing is a policing method or model that
will therefore probably be quite strong, but it doesn't offer tries to forecast, prevent, and minimize crime, have an
any valuable information for forecasting. effective reaction to crime, and identify feasible policing
 Underemphasizing Assessment and Evaluation: Follow- operations, according to the research literature. The use of
up information on both crime incidences and police predictive analytics is a smart strategy to achieve this.
responses must be gathered in order to evaluate the Gathering pertinent and correct data, appropriately
efficacy of crime analysis suggestions. With the processing it in a computer model, and then carrying out a
knowledge that police reactions to the forecasts may have good analysis for the direction and allocation of resources
an impact on the outcomes, the predictive models can be are the steps involved in becoming ready for upcoming
tested on these outcomes over time. For instance, follow- shifts. Advanced and sophisticated algorithms supported by
on analysis can be performed to determine the computer technology are used to carry out this operation.
effectiveness of an intervention in a specific area after a The mathematical model's accuracy as a representation of
hot spot has been discovered and a solution has been reality and its internal consistency are its key
implemented. Police resources may be better allocated bottlenecks.However, due to a natural inability to
elsewhere if the hot spot has been successfully repressed understand statistical probability, the intuitive predictions
for any reason. The reason the hot spot was suppressed, may be prejudiced. Instead of doing this, the brain has
such as a modification in the circumstances or evolved to use holistic associations. Which move quickly
environment, the arrest of a repeat or career offender, or and with little effort. When police personnel are
another factor, may have an impact on the form of the disproportionately exposed to several sorts of crime, the
readjustment. Alternately, if the hot area continues, a availability heuristic that creates a distorted perception of
different intervention should be thought about, and more reality can also affect intuition. This could result in poor
assessments will be required. It's also vital to assess and choices. On the other side, this can be enhanced and
evaluate assertions made about approaches and debiased if police officers assume a conversation with an
software.A vendor might assert that once its software was analyst. Police officers with experience and knowledge do
used, crime in a city decreased by x%. This remark might exist, and to some extent, they can offer appropriate
have been accurate even if the city had not employed the guidance.
program due to larger trends in crime. Any claim of these
tools' usefulness should be viewed with a fair dose of
skepticism in the absence of adequate evaluations.

ISSN No:-2456-2165
REFERENCES Application of big data for national security (pp. 238–
249). Oxford, UK: Butterworth-Heinemann.
[1.] Andrejevic, M. 2017. Digital Citizenship and [18.] Dario, L.M. Morrow, W. J. A Wooditch, Vickovic, S.
Surveillance|To Preempt A Thief International G. 2015. The point break effect: an examination of
Journal of Communication 11: 879. surf, crime, and transitory opportunities. Criminal
[2.] Angwin, J., Larson, J., Mattu, S., and Kirchner, L. justice studies 28 (3), 257-279
2016. ‘Machine bias, Pro Publica. Available [19.] Datta, A., Sen, S., & Zick, Y. 2016, May).
at:https://www.propublica.org/article/machine-bias- Algorithmic transparency via quantitative input
risk-assessments-in-criminal-sentencing. (Accessed: influence: Theory and experiments with learning
25 April 2023). systems. 2016 IEEE Symposium on Security and
[3.] Bachner, J. 2013. Predictive policing: preventing Privacy, 598–617.
crime with data and analytics. Washington, DC: IBM https://dx.doi.org/10.1109/SP.2016.42
Center for the Business of Government. [20.] Drenth, A. (2017) Ervaringen van straatagenten met
[4.] Beck, C, & McCue, C, 2009.“Predictive Policing: het Criminaliteits Edwards, L. and Urquhart, L. 2016.
What Can We Learn from Wal- Mart and Amazon “Privacy in Public Spaces: What Expectations of
about Fighting Crime in a Recession?” Police Chief Privacy Do We Have in Social Media Intelligence?”
Magazine, November 2009. International Journal of Law and Information
[5.] Belina, B. 2016. Predictive policing. Monatsschrift Technology. 24(3): 279-310. DOI:
Kriminol Steuerstrafr 99 (2):85–100’. 10.1093/ijlit/eaw007
[6.] Bennett Moses, L. & Chan, J. 2018, ‘Algorithmic [21.] Egbert, S. & Leese, M. 2021. Criminal futures:
Prediction in Policing: Assumptions, Evaluation, and Predictive policing and everyday police work. Taylor
Accountability ’, Policing and Society, 28: 806–822. & Francis.
[7.] Bennett, M, L, & Chan, J. 2018. Algorithmic [22.] Egbert, S. 2017. Siegeszug der Algorithmen?
Prediction in Policing: Assumptions, Evaluation, and Predictive policing im deutschsprachigen Raum. Aus
Accountability. Policing and Society 28 (7): 806–822. Polit Zeitgesch 67(32–33):17–23.
[8.] Bill Bratton, 2015. “The Profession: A Memoir of [23.] Egbert, S. 2018. About discursive storylines and
Community, Race, and the Arc of Policing in techno-fixes: the political framing of the
America. A famous top cop brags of his successes — implementation of predictive policing in Germany.
and shows his blind spots. The Washington post: Eur J Secur Res. https://doi.org/10.1007/s41125-017-
Democracy die in the darkness. 0027-3.
[9.] Birks, DanTownsley, MichaelHart, Timothy. 2023. [24.] Egbert, Simon, & Susanne Krasmann. 2019.
Predictive policing in an Australian context: “Predictive Policing: Not Yet, But Soon
Assessing viability and utility. ISBN 9781922478870. Preemptive?” Policing and Society 30(8): 905–919.
[10.] Bowers, K. J, Johnson, Shane, D. 2004. Who [25.] Emily Thomas, Why Oakland Police Turned Down
commits near repeats? A test of the boost explanation. Predictive Policing, VICE MOTHERBOARD (Mei.
West Criminol Rev 5(3):12. 28. 2023),
[11.] Braga, A. A, Weisburd, D. 2010. Policing problem http://motherboard.vice.com/en_ca/read/minority-
places: crime hot spots and effective prevention. retort-why-oakland-police-turned-down-predictive-
Oxford University Press, New York. policing (discussing how predictive policing shifted
[12.] Braga, A. A., & Bond, B. J. (2008). Policing crime the strategic focus from generating arrests to
and disorder hot spots: A randomized controlled preventing victimization).
trial. Criminology, 46 (3), 577-607. [26.] Ferguson, A. 2017. "Policing Predictive Policing".
[13.] Brayne, S. 2017. Big Data Surveillance: The Case of Washington University Law Review, 94 (5), 1109–
Policing. American Sociological Review, 82(5), ss. 1189. Available at:
977–1008, doi: 10.1177/0003122417725865. https://openscholarship.wustl.edu/law_lawreview/vol
[14.] Brayne, S. Rosenblat, A. & Boyd, D. in Data & 94/iss5/5.
Civil Rights: A New Era of Policing and Justice [27.] Ferguson, A. G. 2015. Big data and predictive
(2015); available at http://go.nature.com/2j6cot5. reasonable suspicion. University of Pennsylvania Law
Bureau of Justice Statistics. 2009. Law enforcement Review, 63: 327–410.
statistics. Retrieved March 3, 2023, from [28.] Ferguson, A. G. 2019. Predictive policing theory. In
http://www.ojp.usdoj.gov/bjs/lawenf.htm. T.R. Lave and E. J. Miller (eds.), The Cambridge
[15.] Clarke, R. V, Newman, G. R. 2006. Outsmarting the Handbook of Policing in the United States.
terrorists. Präger International, Westport. Cambridge: Cambridge University Press, pp. 491–
[16.] Clarke, R. V., & Weisburd, D. 1994. Diffusion of 510. doi: 10.1017/9781108354721.025.
Crime Control Benefits: Observations on the Reverse [29.] Foucault, M. (1977) Discipline and Punish: The Birth
of Displacement. In Crime Prevention Studies, ed. of the Prison. London: Allen Lane.
Ronald V. Clarke, Vol. 2. Monsey: Criminal Justice [30.] Ferguson, Andrew Guthrie. 2017. The Rise of Big
Press. Data Policing: Surveillance, Race, and the Future of
[17.] Costanzo, P., D’Onofrio, F., & Friedl, J. 2015. Big Law Enforcement. New York: New York University
data and the Italian legal framework: Opportunities Press.
for police forces. In B. Akhgar, G. Saathoff, H. R. [31.] Ferguson, Andrew. 2020. Predictive Policing Theory.
Arabnia, R. Hill,A. Staniforth, & P. S. Bayerl (Eds.), Washington College of Law Research 24: 2020–10.

ISSN No:-2456-2165
[32.] Gluba A. 2016. Mehr offene Fragen als Antworten: Risk: Implications for Criminal Decision Making and
was für eine Bewertung des Nutzens von Predictive Deterrence,” Criminology, 49, 1029–1061. [1].
Policing noch zu klären ist. Die Poliz 107(2):53–57. [46.] Lum, K. and Isaac, W. 2016. ‘To predict and serve?’,
Anticipatie Systeem. Het Tijdschrift voor de Politie, Significance, 13(5), pp. 14–19.
79(3/17), 6. [47.] McCue, M., & Parker, A. 2003. Connecting the dots:
[33.] Groff, E.R. and LaVigne, N.G. 2002. Forecasting the datamining and predictive analytics in law
Future of Predictive Crime Mapping, in Tilley, N. enforcement and intelligence analysis. The Police
(ed.). Analysis for Crime Prevention. Volume. 13. Chief, 10, 10.
Monsey, NY: Criminal Justice Press. pp. 29–58. [48.] Meijer, A. & Wessels, M. 2019. Predictive Policing:
[34.] Haberman C. P. 2017. “Overlapping Hot Spots? Review of Benefits and Drawbacks. International
Examination of the Spatial Heterogeneity of Hot Journal of Public Administration, 42(12), 1031-1039.
Spots of Different Crime Type.” Criminology & [49.] Meijer, Albert, & Martijn Wessels. 2019. Predictive
Public Policy 16:633–60. Policing; Review of Benefits and drawbacks.
[35.] Hardyns, W.Rummens, A. 2018. Predictive policing International Journal of Public Administration 42:
as a new tool for law enforcement? Recent 1031–39.
developments and challenges. European Journal on [50.] Mohler, G. O. Short, M. B. Malinowski, S., Johnson,
Criminal Policy and Research. 2018;24(3):201–218. M. Tita, G. E., Bertozzi, A. L. & Brantingham, P. J.
doi: 10.1007/s10610-017-9361-2. 2015. Randomized controlled field trials of predictive
[36.] Hindelang, M., Gottfredson, M. R. and Garofalo, J. policing. Journal of the American Statistical
1978. Victims of Personal Crime: An Empirical Association, 1459 (November), 00–00.
Foundation for a Theory of Personal Victimization. doi:10.1080/01621459.2015.1077710.
Cambridge, MA: Ballinger. [51.] Mohler, G. O., M. B. Short, P. J. Brantingham, F. P.
https://doi.org/10.1016/j.aci.2016.01.002. Schoenberg, and G. E. Tita, 2011. “Self-Exciting
[37.] Iomio, R., Tears, R., Meadows, L., Becton, J., & Point Process Modeling of Crime,” Journal of the
Charles, M. 2007. The police view of biased-based American Statistical Association, Vol. 106, No. 493,
policing. Police Quarterly, 10(3), 270-287. 2011, pp. 100–108.
[38.] Johnson, S. D., & Bowers, K. J. 2004. The burglary [52.] Mohler, G. O., M. B. Short, Sean Malinowski, Mark
as clue to the future: The beginnings of prospective Johnson, G. E. Tita, Andrea L. Bertozzi, and P. J.
hot spotting. European Journal of Criminology,1(2), Brantingham. 2015. Randomized Controlled Field
237–255. Trials of Predictive Policing. Journal of the American
[39.] Kaufmann, M. Egbert S. & Leese, M. 2018. Statistical Association 110 (512): 1399–1411.
Predictive Policing and the Politics of Patterns ». The [53.] National Crime Prevention Council. 1997. Designing
British Journal of Criminology, safer communities: Crime
https://doi.org/10.1093/bjc/azy060. [54.] Pearsall, B. 2010. Predictive policing: The future of
[40.] Knobloch, T. 2018. Vor Die Lage Kommen: law enforcement. National Institute of Justice
Predictive Policing in Deutschland: Chancen und Journal, 266(1), 16-19.
Gefahren datananalytischer Prognosetechnik und [55.] Pearsall, B. 2019. Predictive Policing: The Future of
Empfehlung Fur den Einsatz in der Polizeiarbeit. Law Enforcement? NIJ Journal, Issue 266, pp. 16-19.
Stiftung Neue Verantwortung & Bertelsman Stiftung. (https://nij.ojp.gov/topics/articles/predictive-policing-
https://www.stiftung-nv.de/sites/defa future-law-enforcement).
ult/files/predictive.policing.pdf [56.] Perry, K. M. Stacy, S. E. & Pepper, C. M. 2019.
[41.] Kump, Paul, David Haro Alonso, Yongyi Yang, Masculine gender-related personality traits and
Joseph Candella, Jonathan Lewin, and Miles N. acquired capability for suicide. Death Studies, 1-8.
Wernick. 2016. ‘Measurement of Repeat Effects in https://doi.org/10.1080/07481187.2019.1699206.
Chicago’s Criminal Social Network.’ Applied [57.] Perry, W. L., McInnis, C. C., Price, S., Smith, S., &
Computing and Informatics, 12 (2), 154–60. Hollywood, J. S. 2013. Predictive policing. The role
[42.] Lammerant, H., & de Hert, P. 2016. Predictive of crime forecasting in law enforcement operations.
profiling and its legal limits: Effectiveness gone Santa Monica: RAND Corporation.
forever. In B. van der Sloot, D. Broeders, & E. [58.] Perry, W.L., McInnis, B., Price, C.P., Smith, S.C. &
Schrijvers (Eds.), Exploring the boundaries of big Hollywood, J. S. 2013. Predictive policing. The role
data (Vol. 32, pp. 145-173). Amsterdam University of crime forecasting in law enforcement operations.
Press/WRR. Prevention through environmental design handbook.
[43.] Lamm-Weisel, D. 2001. "Repeat Victimization for Washington, DC: National
Commercial Burglary and Robbery: How Much and [59.] Richardson, R., Schultz, J. and Crawford, K. 2019.
Where?" Presentation at U.S. Department of Justice, ‘Dirty data, bad predictions: How civil rights
National Institute of Justice, Crime Mapping violations impact police data, predictive policing
Research Conference, Dallas, Texas (December). systems, and justice’. SSRN Scholarly Paper ID
[44.] Landeskriminalamt NRW (2018a). Abschlussbericht 3333423. Rochester, NY: Social Science Research
Projekt SKALA. Düsseldorf. Network. Available at:
[45.] Loughran, T. A., Paternoster, R., Piquero, A. R., and https://papers.ssrn.com/abstract=3333423 (Accessed:
Pogarsky, G. 2011, “On Ambiguity in Perceptions of 15. May, 2023).

ISSN No:-2456-2165
[60.] Robinson, D. & Koepke, L. 2016. Stuck in a Pattern: [74.] van Dijk, B. van Egmond, P. & Hoekstra, M. 2015.
Early evidence on "predictive policing" and civil The Best of Three Worlds in Noord-Holland -
rights. Upturn. Geleerde lessen bij de aanpak van woninginbraken.
[61.] Saunders, J., Hunt, P. & Hollywood, J. S. 2016. Presentation by authors from DSP-Groep and
Predictions put into practice: A quasi-experimental National Police eenheid Noord-Holland.
evaluation of Chicago’s predictive policing pilot. http://www.ctcholland.nl/userfiles/file/18bdB3W_Th
Journal of Experimental Criminology, 12(3), 347– e_Best_of_Three_Worlds_in_Noord- Holland_DSP-
371. doi: 10.1007/s11292-016-9272-0. groep_2015.pdf.
[62.] Schlehahn, E. Aichroth, P. Mann, S., Schreiner, R., [75.] Weisburd, D. 2008. “Place-Based Policing,” Ideas in
Lang, U., Shepherd, D., & Wong, B. 2015. Benefits American Policing, 9, 1–15. [1].
and pitfalls of predictive policing. 2015 European [76.] Weisburd, D. L. & Eck, J. 2004. What Can Police Do
Intelligence and Security Informatics Conference. to Reduce Crime, Disorder, and Fear? The ANNALS
https://dx.doi.org/10.1109/eisic.2015.29. of the American Academy of Political and Social
[63.] Schwabe, C. A. & Schurink, W. J. 2000. Crime cience, 593, 42-65.
forecasting and multivariate analysis at a strategic https://doi.org/10.1177/0002716203262548
level: towards a social science interpretation. [77.] Weisburd, D., Mastrofski, S. D., McNally, A. M.,
Research report for South African Department of Greenspan, R., & Willis, J. J. 2003. Reforming to
Arts, Culture, Science and Technology: Pretoria. preserve: COMPSTAT and strategic problem-solving
[64.] Shapiro, A. 2019. Predictive Policing for Reform? in American policing. Criminology and Public
Indeterminacy and Intervention in Big Data Policing. Policy, 2: 421–456.
Surveillance & Society, 17(3-4): 456-472. [78.] Willems, D. 2014. CAS: Crime anticipation system:
[65.] Shapiro, Aaron. 2017. The Medium is the Mob. predictive policing in Amsterdam. Available at
Media, Culture & Society 39 (6): 930–41. https://event.cwi.nl/mtw2014/media/files/Willems
[66.] Sherman, L. & Weisburd, D. 1995. General deterrent ,%20Dick%20-
effects of police patrol in crime “hot spots”: a %20CAS%20Crime%20anticipation%20system%20_
randomized, controlled trial. Justice Quarterly, 12(4), %20 redicting%20policing%20in%20Amsterdam.pdf
625-648. (last accessed 3 January
[67.] Sherman, L. W, Weisburd, D. 1995. General deterrent [79.] Willems, D., & Doeleman, R. 2014. Predictive
effects of police patrol in crime “hot spots”: a policing – wens of werkelijkheid? Tijdschrift voor de
randomized, controlled trial. Justice Q 12(4):625– Politie, 76(4/5): 39–42.
648. [80.] Williams, M. L., Burnap, P. & Sloan, L. 2016. Crime
[68.] Sommerer, Lucia M. 2017. Geospatial Predictive sensing with big data: the affordances and limitations
policing: Research outlook & a call for legal debate. of using open source communications to estimate
N K Neue Kriminalpolitik 29: 147–64. crime patterns. British Journal of Criminology, 57,
[69.] Sparks, R. F., Glen, A. G., and Dodd, D. J. 1977. 320-340
Surveying victims. New York: Wiley. [81.] Wyant, M. C., C. S. Bretherton, P. N. Blossey, & M.
[70.] The 2016 RAND study of Chicago’s Heat List proves Khairoutdinov (2012), Fast cloud adjustment to
the point. See supra note 192 and accompanying text. increasing CO2 in a superparameterized climate
RAND found the algorithm and program provided no model, J. Adv. Model. Earth Syst., 4, M05001,
statistical evidence that the Heat List worked to doi:10.1029/2011MS000092.
reduce violence. The Chicago Police discounted this [82.] Završnik, A. 2017. Big Data, Crime and Social
criticism, stating that the algorithm had since been Control: Routledge.
improved, so any criticism was dated and https://doi.org/10.4324/9781315395784.
unwarranted. See Stroud, supra note 192.
[71.] Uchida, C. D. 2014. Predictive policing. In: Bruinsma
G, Weisburd D (eds) Encyclopedia of criminology
and criminal justice. Springer, New York, pp 3871–
3880.
[72.] Uchida, Craig. D., 2010. A National Discussion of
Predictive Policing: Defining Our Terms and
Mapping Successful Implementation Strategies,
prepared for the National Institute of Justice,
Document No. NCJ 230404, May 2010. As of May 6,
2023:
https://www.ncjrs.gov/pdffiles1/nij/grants/230404.pdf
[73.] Van Brakel, R. E / Een reflectie over het huidig
toezicht van het gebruik van surveilancetechnologie
door de lokale politie in belgië Cahiers Politiestudies,
volume 55, p. 139 - 160 Posted: 2020.

Key Elements of Predictive Policing Include Crime Analysis, Crime Mapping and Geographies

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Key Elements of Predictive Policing Include Crime Analysis, Crime Mapping and Geographies

Uploaded by

Copyright:

Available Formats

Volume 8, Issue 5, May 2023 International Journal of Innovative Science and Research Technology

Key Elements of Predictive Policing Include Crime

I. INTRODUCTION In reality, the bulk of hotspot and place-based

IJISRT23MAY2191 www.ijisrt.com 3312

IJISRT23MAY2191 www.ijisrt.com 3313

IJISRT23MAY2191 www.ijisrt.com 3314

IJISRT23MAY2191 www.ijisrt.com 3315

IJISRT23MAY2191 www.ijisrt.com 3316

Fig. 1: Prediction-led process (Perry et al., 2013)

IJISRT23MAY2191 www.ijisrt.com 3317

IJISRT23MAY2191 www.ijisrt.com 3318

IJISRT23MAY2191 www.ijisrt.com 3319

IJISRT23MAY2191 www.ijisrt.com 3320

IJISRT23MAY2191 www.ijisrt.com 3321

IJISRT23MAY2191 www.ijisrt.com 3322

IJISRT23MAY2191 www.ijisrt.com 3323

IJISRT23MAY2191 www.ijisrt.com 3324

IJISRT23MAY2191 www.ijisrt.com 3325

IJISRT23MAY2191 www.ijisrt.com 3326

You might also like