C O R P O R AT I O N
Using Future Internet Technologies
to Strengthen Criminal Justice
John S. Hollywood, Dulani Woods, Richard Silberglitt, Brian A. Jackson
K e y findings
• Panelists saw the need to leverage web technologies to improve information-sharing and protection
across the criminal justice enterprise.
• In addition to leveraging web technologies for
information-sharing in general, top priorities
included developing a common criminal history
record and cataloging scheme, developing realtime language translation capabilities, and developing displays or “dashboards” to meet officers’
tailored, dynamic information needs.
• Priorities included general education on key web
technologies, and model policies and procedures
for using them.
• Panelists called for procurement checklists and
cost-benefit tools for systems acquisition, and for
policies and procedures to address the anticipated
S U M M A RY Web technologies just over the horizon,
including semantic tagging, intelligent agents, and the
Internet of Things (IoT), could dramatically change how
the criminal justice enterprise operates. In September
2014, the RAND Corporation convened an expert panel
for the National Institute of Justice to discuss how the
criminal justice community can take advantage of (and
reduce the risks from) these emerging web technologies.
The panel assembled 16 experts on both web technologies and criminal justice, and collectively identified
45 technology needs. After the conference, the panelists
assessed the expected value derived from each need by
rating each need’s potential importance to criminal justice
(law enforcement, courts, or corrections), the technical
feasibility of meeting the need, and the operational feasibility of meeting the need. Top needs from the workshop
are displayed in sidebars. Major themes that cut across
groups of needs are described below.
rise of unmanned vehicles.
• Panelists agreed that the networking infrastructure
needs improvement to support web technologies
(and other applications), especially for courts and
corrections.
• Several needs were expressed related to leveraging wearable and embedded sensors (part of the
Internet of Things), with an emphasis on using sensors to improve officer health and safety.
• Panelists frequently noted the importance of civil
rights, privacy rights, and cybersecurity protections.
Improving Information Sharing
The top theme from the panel was to leverage web technologies to improve information sharing and protection
across the criminal justice enterprise. In addition to leveraging web technologies for information-sharing in general,
top priorities included developing a common criminal history record and cataloging scheme, developing real-time
language translation capabilities, and developing displays
or “dashboards” to meet officers’ tailored, dynamic information needs.
2
Educating Practitioners
Another major theme was improving practitioners’ knowledge of web technologies. Priorities included general
education on key web technologies, and model policies and
procedures for using them. Panelists also called for procurement checklists and cost-benefit tools for systems acquisition, and policies and for procedures to address the anticipated rise of unmanned vehicles.
Improving Infrastructure
A third theme was to improve the networking infrastructure
needed to support web technologies (and other applications), especially for courts and corrections.
Exploring the Use of Emerging Internet of Things
(IoT) Sensors in Criminal Justice
Several needs were expressed related to leveraging wearable and
embedded sensors (part of the IoT), with an emphasis on using
sensors to improve officer health and safety.
Civil Rights, Privacy Rights and Cybersecurity
Concerns
Panelists frequently noted the importance of civil rights, privacy
rights, and cybersecurity protections. While there were few needs
about these topics specifically, panelists noted that more than half
of the needs either raised security, privacy, or civil rights concerns
or had implied requirements on these topics.
Top Five Needs Overall
•
•
•
•
•
Video links to correctional facilities, so inmates can meet
with community corrections officers and others prior to
release
Educational materials on key web technologies (sensors,
video conferencing, tele-education, data translation)
Real-time language interpretation
Virtual criminal record catalog
Technological infrastructure for criminal justice interfaces
Top 5 Needs of Law Enforcement
•
•
•
•
•
Policies and procedures for
unmanned and automated vehicles
Virtual criminal record catalog
Better access to data for facial
recognition identification
Biomedical sensors for officers
Identification of officers in close proximity
Top 5 Needs of Courts
•
•
•
•
•
Video links to correctional facilities
Procurement checklist for courts’
information technology
High-speed Internet connections for courts
Virtual courtrooms
Educational materials on key web
technologies
Top 5 Needs of Corrections
•
•
•
•
•
Video links to correctional facilities
Better access to data for facial
recognition identification
Biomedical sensors for officers
Internet of Things–enabled models for house
arrest
Educational materials on key web
technologies
Conclusions and Recommendations
Information sharing: Partner with the Standards Coordinating
Council and constituent information-sharing development efforts to explore how semantic tags and intelligent agents
might be leveraged to expedite information sharing, with criminal history data as a starting point. Experiment with realtime language technologies.
Practitioners’ knowledge: Focus education efforts on semantic technologies that support finding, accessing, and translating key information; sensor systems for monitoring officer health and officer safety, and maintaining community supervision; video conferencing; and civil rights, privacy rights, and cybersecurity protections.
Designate a group to develop law enforcement requirements, policies, and procedures, for interfacing with self-driving cars.
Infrastructure: Develop field experiments with video teleconferencing links for inmate communications and remote
education. Pursue novel business models and support to make Internet links more affordable in rural areas.
3
IoT sensors: Experiment with health and safety sensor feeds, both wearable and embedded. Experiment with
Internet-connected sensor systems to support maintaining the location and tracking of offenders under community corrections supervision.
Civil rights, privacy rights, and cybersecurity: At a strategic level, seek to ensure that civil rights, privacy rights, and
cybersecurity provisions are built into technology developments, standards, policies, and procedures from the beginning.
For intelligent agents that support decisionmaking, research how to ensure the quality of data used to make the decision,
and how decisionmakers should use the agents’ recommendations. For IoT sensors, conduct research to advise on common
attributes for policies, procedures, and required protective technologies for sensors related to the IoT.
I NTRO DUCTI O N
The police officer directing traffic in the intersection could see
the car barreling toward him and the occupant looking down at his
smartphone. Officer Rodriguez gestured for the car to stop, and the
self-driving vehicle rolled to a halt behind the crosswalk.
The officer waved the car on as the oblivious passenger continued checking his email. But he wasn’t oblivious for long. A very
human-driven sport-utility vehicle (SUV) barreled through the
intersection, forcing the officer to dive for safety and the automated
car to brake hard and swerve to avoid a collision.
While Officer Rodriguez called for assistance, an unmanned
aerial vehicle on patrol recognized the speeding SUV and gave
chase while it transmitted the vehicle’s location to police cruisers. A
police cruiser precision immobilization technique (PIT) maneuver
forced the SUV off the road minutes later.1 As officers prepared
to swarm the vehicle, one took the man’s photo from a distance,
uploaded it to compare against a national repository of mug
shots, and quickly produced a high-probability match. The photo,
combined with the license plate and vehicle description, helped the
system identify an ex-convict who was related to the SUV’s owner,
and who had a lengthy rap sheet of armed robbery and reckless
driving.
A few moments later, an armed robbery at a nearby gas station crackled over the officers’ radios. Surveillance photos from the
scene, showing the SUV driver pointing a weapon at the gas station
attendant, arrived shortly after the suspect had been safely taken
into custody—where he would remain for a very long time.
———
How we use information technology (IT) is on the verge of
dramatic change. In the past, the World Wide Web and the
Internet largely have involved passive activities: reading or
watching multimedia on a computer screen or typing in search
requests on a keyboard or touch screen. Over time, we have
migrated from personal computers to laptops to tablets and
smartphones, but the basic interactions with the web have
been the same. In the future, however, the web will be embedded in our surroundings—in clothes, cars, appliances, flying
unmanned vehicles, and utility grids themselves. Rather than
typing in search requests, information services will be delivered
to us somewhat automatically, depending on context. Criminal
justice agencies must prepare to make the most of this transition while protecting civilians from the threats these new
technologies may present.
The web as we have known it—mostly web pages and
social media—has had a significant impact on criminal justice.
For example, the massive-scale “crowdsourcing” of pictures and
video from the Boston Marathon bombing generated a great
deal of useful intelligence for investigators (Wadwha, 2013).
However, the criminal justice community often is perceived as
largely reacting to new web technologies (and other technologies; see, for example, Smith, 2013) rather than anticipating
them. In addition, not every innovation is perfect or always useful to criminal justice. The same crowdsourcing that produced
useful intelligence after the Boston Marathon bombing also
resulted in a number of innocent people being wrongly accused
(Wadwha [2013] notes that the crowdsourcing was a great success at gathering information but a failure at investigating it).
To help the criminal justice community learn about upcoming web technologies, and to start informing developers and
policymakers about what criminal justice practitioners will need
to use new technologies effectively, the RAND Corporation and
the National Institute of Justice (NIJ) brought together an expert
panel of 16 practitioners and technology experts. Collectively,
they discussed what upcoming web technologies are valuable
and what they likely will mean for criminal justice. Then they
identified and prioritized 45 needs for both technology and
policy development that would allow usage of new technologies
4
successfully while mitigating the technologies’ risks. This report
summarizes the panel’s discussion, first describing what the new
technologies are and how they might be used, then covering the
specific needs for technology and policy development.
EM E RG I N G W E B TEC H N O LO G I ES AN D
C R IM I NAL J USTI C E
The current web technologies that we regularly use are commonly referred to as Web 2.0 technologies. These include basic
web pages that look like documents, searches based on keywords and phrases, online mapping and navigation, and social
media. Table 1 adapts Kevin Kelly’s perspective on the history
of the web and associated information technologies (2007).
While there are no canonical definitions of web generations,
Kelly’s descriptions are some of the most widely disseminated. Kelly characterizes the linking of documents, individuals, and data as generations 1.0, 2.0, and 3.0 of the web. In the
table, we extend his characterization into the past and future by
examining the IT linkages that came before and the ones that
are just over the horizon. As noted, Web 2.0 can be thought of
as today’s web, as social media and crowdsourcing technologies
(along with the Web 1.0 baseline technologies of document
browsing and searching) are ubiquitous. Web 3.0 and 4.0 technologies are emerging now.
The key technologies for Web 3.0 (predominantly Semantic
Web–related) tend to be foundational in nature: They support
the information management and sharing needed to enable the
more immediately visible technologies (intelligent agents and
Internet of Things [IoT]) of Web 4.0.
Web 3.0 and 4.0 technologies will have a variety of potential applications and effects on the criminal justice system. As
These technologies
will improve the ability
of average citizens
and criminal justice
practitioners to “see”
around corners.
a whole, these technologies will change the way citizens obtain
and synthesize information. They will offer cheaper and more
accessible methods for physically monitoring our personal
environments or any other environment of interest. In addition,
they will offer the ability to intelligently and remotely react to
changes in those environments.
Imagining the Technological Future of
Criminal Justice
These technologies will improve the ability of average citizens
and criminal justice practitioners to “see” around corners. This
improved sight will be physical in terms of remote monitoring,
and informational in terms of having access to (or being intelligently supplied with) highly relevant, just-in-time information
about individuals’ identities, connections, reputations, histories
(including criminal), past and present whereabouts, etc.
Imagine a law enforcement officer interacting with a
vehicle that has sensors connected to the Internet. With the
appropriate judicial clearances, an officer could ask the vehicle
to identify its occupants and location histories. The officer then
could use Semantic Web technologies to review the criminal
histories and search for any outstanding warrants of the occupants across dozens or even hundreds of local, state, and federal
repositories—even repositories that do not contain data in a
traditional “compatible data format” (but that are semantically
tagged). Or, if the vehicle is unmanned but capable of autonomous movement and in an undesirable location (for example,
parked illegally or in the immediate vicinity of an emergency),
an officer could direct the vehicle to move to a new location
(with the vehicle’s intelligent agents recognizing “officer” and
“directions to move”) and automatically notify its owner and
occupants.
Individuals on pretrial or postconviction supervision are
already being fitted with location monitoring devices. As the
communication and processing capabilities of these devices
improve, they also might monitor these individuals’ stress
levels, drug use, or states of arousal (for sex offenders). They
could allow for video or telephonic connections to monitored
individuals and provide the monitoring officer with clues about
the truthfulness of verbal responses. More broadly, suppose that
an offender disables a tracking bracelet before leaving his or
her home to commit a crime. Other sensors in the home could
notify police of the offender’s departure.
Of course, all of this “vision,” monitoring, and sensitive
information sharing raises substantial security, privacy, and
5
Table 1. Generations of Linkages in Information Technology
Generation
What Was Linked?
Characteristics
0.0
Nothing (stand-alone personal computers)
Vast computational power becomes accessible to the masses.
0.5
Computers were linked in local networks,
which were further linked into Internetworks.
The capability was developed to share such resources as files, printing,
processing, and storage.
1.0
Documents were linked to create the World
Wide Web. Search engines helped users
discover these documents through keyword
searches.
Initial use was an extension of existing content and information sources
(broadcast, publishing, etc.), with institutions making online content
available to users.
2.0
Individuals and groups are linked to
form social networks. Networks of
individuals are able to collaborate en
masse to intelligently achieve objectives
(crowdsourcing).
Information flow becomes much more collaborative and bidirectional with
sharing of text, images, video, etc. Users become content creators, not
just readers of institutionally published information. Crowdsourcing is used
to identify and locate suspects, identify crime hotspots, and address other
problems in emergency response and beyond.
3.0
Semantic Web: Data are semantically
tagged and linked. Information searches
use context and natural language instead
of just keywords. Researchers have access
to datasets that are linked together through
formal, structured relationships.
Computers are better able to “understand” the content of the information
processed and “communicate” with humans using spoken and written
language. Difficult analytic problems can be solved using data that was
previously difficult to integrate (such as medical records and medical
research). With semantic tagging and linking, one can “ask” the following
types of requests:
• “Find scientific publications like those written by these specified
authors”
• “Find pictures of Georgian chests of drawers for sale near my house”
• “Find other employees whose backgrounds make them a good fit for
my project”
• “Display a network chart of all technical design documents for this
wing assembly and the relationships between the documents”
• “Show me points of interest around me as I walk down this street”
(also relevant to the IoT, below).a
4.0
Intelligent agents: Intelligent software
agents learn and perform rudimentary
tasks.
Software tools learn what sorts of information and analysis a user wants,
and delivers it to him or her regularly. Some of these tools are embedded
in devices (see IoT, below). Two key types of intelligent agents are
especially relevant to criminal justice:
• search agents that scan a range of media (social media postings,
video, photos, etc.) to find events “of interest.” In the criminal justice
context, these tend to relate to suspicious activity or a possible
connection to a crime.
• decision agents that take in a range of data and use statistical
algorithms to advise someone on how to make a decision. In the
criminal justice context, these might include judgments on bail or
sentencing.
Internet of Things: Previously unlinked
everyday objects are now linked to the
Internet: automobile, refrigerator, watch,
smoke detector, thermostat, airline seat, etc.
The services provided by these objects are enhanced. Connected cars
can improve traffic flow. Things become sensors that can collaborate
to improve life. One example is buildings that adapt to weather, warn/
defend against external emergencies (earthquake, tornado, fire, flood,
etc.), or report the number of building occupants, their locations, and vital
signs. Autonomous vehicles can be thought of as part of both intelligent
agents and the IoT.
These questions are adapted in part from examples taken from the World Wide Web Consortium’s draft requirements for an initial web ontology language
(Heflin, Volz, and Dale, 2002).
a
6
Interagency
communication has been
a perennial problem for all
organizations, and criminal
justice organizations are
no different.
civil rights questions; we discuss these issues further below. In
the worst case, individuals’ private behavior and personal data
could be stolen and exploited, either by criminal hackers or corrupt government officials.
Interagency communication has been a perennial problem
for all organizations, and criminal justice organizations are no
different. Lawyers and courts have a constant need to schedule
appearances efficiently, and law enforcement agencies need up-todate information on the status of parolees when they encounter
them. Many agencies currently rely on Web 1.0 technologies to
address this need. Community corrections officers (whom courts
often employ), periodically export lists of parolees by email or
host them on secure file transfer protocol sites, transferring files
from one host to another for law enforcement use. These are
transmitted in “standard” formats, such as extensible markup
language (XML) or Excel worksheets. Web 3.0 technologies,
such as linked data, will facilitate on-demand access to this
information and make it available to a larger number of agencies. Semantic Web and linked data techniques can facilitate the
interconnection of documents, social networks, and databases,
without regard to the structure of the underlying data source
(provided the data are tagged and stored in a core format that
can be searched, retrieved, and displayed readily). When fully
realized in this context, linked data techniques will provide law
enforcement with the most accurate and timely information
while providing parole officers with important details about
parolees’ contact with law enforcement.
The scheduling of courtrooms, judges, lawyers, and others
in the criminal justice system was noted as a problem during
the workshop. In the future, a court system could use intelligent software agents working on behalf of their human and
physical (courtroom) counterparts to automatically and intelli-
gently examine and prioritize individual schedules and dynamically assemble a court docket. Once assembled, the individual
agents could push updated information to other interested
parties and the public.
It is important to address another technology that is not
actually new: cheap and portable personal video teleconference
equipment (on smartphones, computers, and other devices),
which is nearly ubiquitous. The expert panel discussed this at
length due to its potential uses in the law enforcement, court,
and prison systems. What is new is the growing accessibilityon-demand and interconnection of all these systems, made
possible by emerging Web 3.0 and 4.0 standards. Consequently,
judges might be able to set up court nearly anywhere (“courtroom in a box”) or receive remote testimony from anywhere.
Prisoners might be able to better maintain family connections,
build connections with their parole officer and social service
providers, and generally improve their transition to civilian life
through easy and inexpensive videoconferencing—all while
intelligent agents monitor their communications to flag any
discussions that appear to reference criminal activity.
However, the dark side to all of the emerging access and
interconnectivity is the risk to the public’s civil rights, privacy
rights, and security. One can readily imagine abuses that might
occur if, for example, capabilities to control automated vehicles
and the disclosure of detailed personal information about their
occupants were not tightly controlled and secured. Intelligent
agents monitoring social media feeds might wrongly flag certain people as potential suspects—notwithstanding larger issues
of the circumstances under which such monitoring is justified.
Policies, procedures, and technologies to help ensure that new
technologies are not abused were a major discussion throughout
the workshop.
Methodology
To consider the implications of these rapidly emerging technologies for criminal justice—as well as how the criminal
justice community might get ahead of the curve in using
them—NIJ asked RAND to assemble an expert panel of both
criminal justice practitioners and technologists. In all, 16 panelists participated; see the appendix for the complete list. On the
practitioner side, we asked the major practitioner associations in
law enforcement, courts, and corrections to send representatives
to serve as the practitioner experts. Six associations sent representatives: the International Association of Crime Analysts,
the International Association of Chiefs of Police, the American
7
Jail Association, the American Probation and Parole Association, the American Correctional Association, and the National
Center for State Courts.
On the technologist side, RAND identified a list of experts
with substantial backgrounds in the key web technologies
under consideration (intelligent agents, Semantic Web, and
sensors/IoT), as well as general expertise with IT in the public
sector. RAND Knowledge Services, in particular, conducted a
detailed search of experts’ online information to prepare brief
profiles of candidates, with those who had some experience
working in the public sector (especially security and/or criminal
justice) given priority for potential participation. Ten of these
technical experts agreed to attend the workshop.
In hosting the panel, we recognized that since Web 3.0+
technologies are new to criminal justice, few attendees would
have experience both with the new technologies and with
criminal justice technologies and practices. Thus, our concept
was to bring together those with both web technology expertise
and criminal justice expertise and provide orientation materials
and sessions to share information about both web technologies
trends and criminal justice technology perspectives. We then
had the panelists work together in small breakout groups to
jointly discuss pressing problems in criminal justice and how
the new technologies might help or hinder solutions, as well as
general opportunities and challenges the new web technologies
might raise for criminal justice. We then had the panelists use
these issues to generate specific needs to leverage the opportunities and mitigate the challenges posed by Web 3.0+ technologies; those needs could include technical development, policy
changes, practices changes, and training changes.
After departing the workshop, the panel convened electronically in multiple rounds to consolidate, edit, categorize,
and prioritize the needs. The technical details are described in
the appendix to this report. In brief, panelists rated how valu-
able a solution to each need might be to each community of
practice (law enforcement, courts, and corrections). They then
assessed how technically feasible and operationally feasible it
would be to develop and field those solutions. (Here, panelists were asked to explicitly include affordability, civil rights,
privacy rights, and security concerns in making operational
feasibility ratings.) We then combined these ratings to produce
expected value (EV) scores—how much value a solution might
provide multiplied by the likelihood a solution actually could
be produced and fielded.
To provide rough assessments of the needs priorities, we used
a clustering tool to divide the needs into three tiers: Tier 1 (high
priority), Tier 2 (medium priority), and Tier 3 (low priority). The
tool was used to find the “best overall” splits to divide the needs
into tiers, with “best” measured in a mathematical sense.
The panelists frequently flagged particular needs as raising
civil rights, privacy rights, or cybersecurity issues. We tracked
which needs were flagged as raising one or more of these issues
as well.
We believe that we assembled an experienced group of
panelists who produced a well-founded set of technology needs.
The final panel was split roughly 50–50 between those who
had web technologies expertise and those who had expertise on
criminal justice IT and practices. Panel members’ specific areas
of expertise included
• criminal justice technologies, standards, and practices,
focusing on IT for institutional corrections, community
corrections, courts, crime analysis, and law enforcement
• Semantic Web technology, to include providing education about Semantic Web technology in general and using
semantic links to improve what is known about the provenance and trustworthiness of data
Our concept was to bring together those with both
web technology expertise and criminal justice expertise
and provide orientation materials and sessions to share
information about both web technologies trends and
criminal justice technology perspectives.
8
• the IoT, including developing both hardware and software
(on-device and backend processing)
• public sector IT, including experience with eGovernment
and Open Government initiatives, serving as a state and
agency chief information officer, and expertise in communications and computing infrastructure providing highbandwidth data and video
• technologies related to Web 3.0+, including IT security,
civil rights, and digital privacy issues and solutions; security and management controls and audits specifically for
government agencies; XML and transaction processing;
and geospatial analysis, including recognizing, geocoding,
and visualizing event location data with assorted details.
However, we recognize that these results reflect their subjective assessments, and as with any panel exercise involving a
limited number of participants, a different group may well have
produced a different set of results. That said, our findings are
consistent with our prior studies on criminal justice technology
needs (Hollywood et al., 2015, Jackson et al., 2015, Silberglitt
et al., 2015) and with findings from earlier studies on law
enforcement technology needs (Koper, Taylor, and Kubu, 2009;
International Association of Chiefs of Police, 2005), and we
believe they will provide a useful guide to both practitioners and
developers in considering how to leverage Web 3.0+ technologies.
Top Technology Needs from the Workshop
for Each Community of Practice
Table 2 summarizes the Tier 1 needs overall and for each criminal justice community of practice. The needs are shown in order
of their expected value for that community of practice, with the
need with the highest score shown first.
One need has a single star—providing video links to correctional facilities had scores noticeably higher than the scores for
all other needs. Two needs at the bottom of the corrections list
are double-starred. These are needs that were initially assigned to
Tier 2 by the predictive analytics tool but had noticeably higher
scores than other Tier 2 needs. They are included in Table 2
because there are comparatively few Tier 1 needs for corrections
as a percentage of all corrections needs (there are fewer Tier 1
needs for courts, but there also were fewer courts needs overall).
Figure 1 summarizes the needs by technology area. As
shown, three-quarters of the needs are related to intelligent
agents (both to search for suspicious activity and to help personnel make decisions in the field), standards leveraging the
Semantic Web, policies and procedures for using the emerging
technologies, and sensors in the IoT. The remaining quarter
of the needs under “Other” covers a wide range of technical
topics. These included improving networking infrastructure,
research on displaying information, real-time language interpretation, real-time decryption of criminals’ codes, facial recognition, identifying structured data in free text, and interfaces
with unmanned vehicles.
On average, the scores for search agents were on the low
side, whereas the scores for standards and policies, procedures,
and guidance were on the high side. The lower scores for search
agents appear due in part to feasibility issues (such as whether
search agents can really find genuinely worrisome activity going
on while avoiding “finding” huge numbers of innocent activities).
The lower scores also were due in part to concerns about civil and
privacy rights (since, depending on the application, many agents
would be searching through large amounts of public data). The
higher scores for standards and policies, procedures, and guidance appear to be due to (1) being seen as necessary to the use
of web technologies across the criminal justice community and
(2) being comparatively easy to develop and implement.
D ISCUSSI O N
In prior research on priority criminal justice technology needs
(International Association of Chiefs of Police, 2005, Koper,
Taylor, and Kubu, 2009, and Hollywood et al., 2015, on needs
Figure 1. Needs by Technology Category
Other
24%
IoT sensors
(wearable/
embedded)
13%
Policies,
procedures,
and guidance
13%
RAND RR928-1
Search agents
18%
Decision agents
16%
Standards
16%
9
Table 2. Top Needs Overall and for Each Community of Practice
Overall
Video links to correctional facilities*
Educational materials on key web technologies
Real-time language interpretation
Virtual criminal record catalog
Technological infrastructure for criminal justice interfaces (elements needed for information sharing)
Funding high-speed Internet connections for courts
Tools to improve data quality and integrity
Cost-benefit tools for technology acquisition
Policies and procedures for emerging IoT and Semantic Web technologies
Improved sharing of information about offenders with third parties (treatment providers and other stakeholders)
Better access to data for facial recognition identification
Biomedical sensors for officers
Research on information overload
Intelligent agents to help protect digital evidence chain of custody
Situational awareness/mapping displays tailored to individual officers
Improved tracking of officers within buildings
Procurement checklist for courts IT
Law Enforcement
Policies and procedures for unmanned and automated vehicles
Virtual criminal record catalog
Better access to data for facial recognition identification
Biomedical sensors for officers
Identification of officers in close proximity
Educational materials on key web technologies
Research on information overload
Video links to facilities
Real-time language interpretation
Situational awareness/mapping displays tailored to individual officers
Analytics for social media and community feedback
Improved tracking of officers within buildings
Tool to assess skills of individuals at an incident
Policies and procedures for emerging IoT and Semantic Web technologies
Corrections
Video links to correctional facilities*
Better access to data for facial recognition identification
Biomedical sensors for officers
IoT-enabled models for house arrest
Educational materials on key web technologies
High-speed Internet connections for courts
Situational awareness/mapping displays tailored to individual officers**
Improved tracking of officers within buildings**
Courts
Video links to correctional facilities
Procurement checklist for courts IT
High-speed Internet connections for courts
Virtual courtrooms
Educational materials on key web technologies
Real-time language interpretation
* Scores noticeably higher than scores for all other needs.
** Initially assigned to Tier 2 by the predictive analytics tool but had noticeably higher scores than other Tier 2 needs
10
The call for “technological infrastructure” closely matches
the general desire to improve criminal justice information
sharing, in a single need.
for law enforcement; Jackson et al., 2015, on needs for corrections; and Silberglitt et al., 2015, on future needs for law
enforcement), top needs could be grouped into three overarching themes. The first reflects a general demand for greater
knowledge and educational development about technologies
as well as how to apply them effectively. The second reflects a
general demand to improve the sharing and display of information as needed across the criminal justice enterprise. The
third comprises “everything else.” That same pattern applies to
the results from the Web 3.0+ technologies workshop, with a
few additional themes emerging: a need to improve networking infrastructure; a need to develop IoT sensors for criminal
justice; and a need to ensure protections of civil rights, privacy
right, and cybersecurity around the use of web technologies in
criminal justice. The specifics are discussed below.
Theme 1: Improving Information Sharing
The driver stopped by Officer Nguyen appeared to be completely
confused by the officer’s request to produce his license and registration. “No entiendo,” he said. “No hablo Inglés.” The officer’s earpiece
immediately translated: “I don’t understand. I don’t speak English.”
Officer Nguyen held out a small speaker and spoke quietly
into a microphone. In Spanish, the man heard the officer’s words:
“That’s okay, I just need your license and registration.”
———
The cluster of demands for improved information sharing had
the highest number of top-tier needs (eight). These are, in order
of overall expected value (with the most needed at the top),
• real-time language interpretation (ranked third overall)
• virtual criminal record catalog (ranked second for law
enforcement and fourth overall)
• technological infrastructure for criminal justice interfaces
(ranked fifth overall)
• tools to improve data quality and integrity
• improved sharing of information about offenders with
third parties (treatment providers and other stakeholders)
• better access to data for facial recognition identification
(ranked second for corrections)
• research on information overload
• situational awareness/mapping displays tailored to individual officers.
All of these demands were discussed in the context of leveraging web technologies, especially Semantic Web (3.0 technologies), to support meeting them. We discuss how web technologies might be leveraged to improve information sharing and
security below, first in the general sense, and then with regard
to the specific needs under this theme.
Web Technologies to Improve Information Sharing
and Safeguarding in General
Panelists thought it would be useful to explore whether semantic tagging might help facilitate data sharing in general, along
with a range of data safeguards (on data provenance, privacy,
and data access) and whether intelligent agents might be used
to monitor and enforce access control and usage policies. As an
initial focus, panelists thought it would be useful to develop
standards for criminal history data, to include sharing specifications, cataloging specifications, and common policies.
The call for “technological infrastructure” closely matches
the general desire to improve criminal justice information
sharing, in a single need. Hollywood et al. (2015) touch on the
substantial difficulties and complexities involved in developing
all the components needed for genuine, seamless information
sharing, while acknowledging the substantial progress made to
date. During the workshop, there were discussions on how web
technologies, especially semantic tagging, with some assistance
from intelligent agents, might facilitate sharing and safeguarding information, in general. (Here, “safeguarding” includes
both cybersecurity and protecting the quality and integrity of
the information.)
To get a sense of how semantic and agent technologies
might help, we discuss the major advance between Web 1.0/2.0
11
and Web 3.0 in more detail. Coding (tagging) of web pages in
Web 1.0 and 2.0 was primarily about formatting how the pages
would appear—where to make text appear in bold or in italics,
what color to make the page, where and how to put content
in tables, and so on. Relationships were included, but just on
the level of “click here to go this other web page.” While one
certainly could—and did—support searching for information
across pages, the searching was based on matching key words
and terms, with Google and others adding methods for guessing which pages a user would likely want, based on which pages
had the most inbound links, where they were searching from,
and so on (Page et al., 1998, provides an early discussion of
Google’s algorithms, for example).
We also note, today, that most criminal justice data are
stored in relational databases. This means data are stored in
large structured tables, with records as the rows of the table
and fields as the columns. Importantly, each table entry typically is a small piece of structured data, such as a number or a
small text string (“street name, no more than 80 characters”).
Each record has a unique identifier, called a key. In a records
management system, for example, one typically has data
tables describing basic facts about persons, a second describing
locations, a third describing tables, and other tables describing facts about incidents. (See, for example, Law Enforcement
Information Technology Standard Council, 2008). Pulling
a person’s criminal history involves querying and joining
information both about the person (from a “names table”)
and about the past crimes they committed (from the incidents
table). The difficulty is compounded if one wants to pull a
person’s criminal history across multiple jurisdictions where
he or she committed crimes—one has to put in very specific
queries to a whole range of disjointed databases in different
agencies, and then be able to properly interpret and assemble
the results. Consider how different agencies’ databases might
treat “name” differently: Is it first name/last name? Is it
all one string? Are the middle initial or full middle name
included, and if so, where? What if someone is a Sr., Jr., or
III? And these questions are just for a person’s name. It is easy
to see how databases are naturally noninteroperable with each other.
What the Semantic Web does is add codes (tags) that
characterize the content of information in a web page, as well
as relationships to related information, not just describe how to
display the information. For example, a person can now have
a page of machine-readable data about themselves that looks
like:
John Smith
Despite the symbols (tags), it is clear that we have described
a person named John Smith along with his homepage and a
picture of him. The tags make it possible for computer software
agents to determine that as well. It is similarly possible to add
machine-readable fields describing organizations to which John
Smith belongs, who his associates are (linking to their machinereadable profiles), and where he lives and works. The semantic
tags shown are part of the Friend-of-a-Friend (FOAF) format
(Foaf Project, undated).
New types of databases also are being deployed, with a key
type being documented-oriented databases (Lerman, 2011).
Rather than being a set of small-tightly-controlled fields in a
table, each record is now a text document filled with data coded
with tags such as the one shown above, as well as contenttagged links to other related information, which can be inside
or outside of the database. The format makes it very easy to add
new information—if a subject is convicted of another crime, for
example, his or her criminal history document can be updated
with that conviction event by adding a few lines of tagged text.
We can see how the use of these technologies might make
sharing criminal justice information much easier. Suppose
an officer needs to look up the criminal history of a suspect
quickly:
• First, it will be much easier to search specifically for criminal history records across a range of systems, since they will
be specifically tagged as criminal history records.
• Finding the right record will also be much easier, since the
officer will be able to search specifically on what is known
New types of databases
also are being deployed,
with a key type being
documented-oriented
databases.
12
about the person—name, address, scars/marks/tattoos, or
any other specific information known about the person.
• It will be much easier to put together the criminal history
of the subject, as almost all of it will be part of one document record, perhaps with links to related information
(photos, detailed reports about specific incidents, etc.). The
history will not have to be assembled across multiple tables
and databases.
• Finally, it will be much easier to look up and review the
related, linked information, as the linked information itself
will be expressed in terms of these tagged text documents.
It is much easier and much more robust to interpret the
tagged text than to have to deal with the very specific
file and data formats used within traditional databases.
For example, we will know a block of text in a document
record is an as opposed to having to interpret
how an agency’s internal database chops up and represents
all of the different parts of an address.
Intelligent agents, especially search agents, hold the promise of
leveraging these tags to simplify and automate the searching
and retrieval described above.
New web technologies developments have promise for
making information safeguarding much easier as well. The
same sorts of tags described above can also be used to describe
the security requirements to access a document (or a database of
documents). Identity and access management systems for users
similarly have tagged documents describing information about
each user, so in principle, one can then match up the users’
credentials with document access requirements to determine
whether a given user can get a given document. The goal is to
move toward “single sign-on” systems in which a user needs to
log into only one portal to get access to all of the information
New web technologies
developments have
promise for making
information safeguarding
much easier as well.
needed to do his or her job, as opposed to having to log in to up
to dozens of separate systems.
Further, it is now becoming possible to largely automate
one of the most painful parts of information sharing—negotiating memoranda of understanding between multiple agencies to share information. These depend in large part on all
parties agreeing on whom (more specifically, which roles) will
get access to the information and what security measures
will be taken on both ends with the data. In an initiative
called Trustmarks, both those seeking information and those
providing information can get certified tags (the trustmarks)
describing their roles and the security measures with which
they comply (as well as the auditing mechanisms checking
compliance), making it possible to automatically identify
whether two agencies can share based on matching roles and
trustmarks (GTRI NSTIC Trustmark Pilot, undated). As
with searching, intelligent agents can simplify and automate
the process of interpreting trustmarks and other security tags
to get access to needed information.
Opportunities for and Obstacles to Improving
Information Sharing in General
Both the biggest opportunity for and obstacle to information
sharing is the large number of information-sharing initiatives
under way. A top finding from Hollywood et al. (2015) was
that integration and dissemination of existing informationsharing efforts were needed far more than new starts. To
incorporate Semantic Web and other technologies into
criminal justice information sharing would involve coordinating Global Justice Information Sharing Initiative (GLOBAL),
National Information Exchange Model (NIEM), and other
key developers (Federal Bureau of Investigation Criminal
Justice Information Systems, IJIS Institute standards and
testing development,2 and International Association of Chiefs
of Police Model Policies are a few key examples) to determine
whether web technologies might help expedite building information exchange and information assurance standards. The
recently formed Standards Coordination Council, an advisory
working group to the White House on information sharing
and safeguarding standards, includes 14 different organizations (Standards Coordinating Council, undated).
The number of different standards themselves is far
larger. For example, NIEM information exchange package
documents (IEPDs) are key tools for supporting information
sharing across multiple systems. However, there are hundreds
13
of such IEPDs, and they overlap and can be inconsistent.3
A search for “criminal history” on the IEPD Clearinghouse
(undated) alone found 91 relevant IEPDs. Coordination and
integration will be one of the biggest challenges. Policies and
procedures for sharing this information with third-party
providers such as treatment providers would address an additional top-priority need. Similarly, developing policies and
procedures for emerging IoT might be folded into existing
efforts toward developing model IT policy.
Specific Needs Related to Information Sharing
Here we discuss top-rated information sharing needs that were
highly specific.
Better access to facial data addresses the specific problem of
not being able to query the full range of federal, state, and local
databases containing mug shots. It was seen as a policy problem
as much as a technical data-exchange problem. This need did
raise substantial concerns about civil rights and privacy rights
that included ensuring proper access to sensitive information
and the risk of false matches (for example, identifying someone
as a criminal based solely on his or her appearance). Panelists
also noted that addressing this need would have to be consistent
with the Federal Bureau of Investigation’s (FBI’s) new Interstate
Photo System (part of the FBI’s larger Next Generation Identification system).
The call for “situational awareness displays” to use the shared
information effectively closely parallels a similar need for law
enforcement IT (Hollywood et al., 2015). Also related is a call
for human factors research on averting information overload
from all the incoming data.
As with information-sharing infrastructure, ongoing
research and development is related to developing situational
awareness displays, or “dashboards,” for staff at different
levels. Semantic tags and intelligent agents might contribute to
these research and development (R&D) efforts. For example,
semantic tags might lead to better ways to display what is most
relevant in various situations while hiding what is not. Simi-
larly, intelligent agents populating the dashboards offer the possibility of addressing information overloads—assuming that the
intelligent agents can be readily trained to give users the types
of information they need when they need it.
The highest-ranking machine-to-user need—and the
highest-ranking information-sharing need overall—was to
develop real-time language interpretation services. There does
appear to be substantial commercial development in this area
that could be leveraged; for example, Microsoft in December
2014 released a real-time English-to-Spanish translation service
for Skype (Warren, 2014).
Theme 2: Improving Practitioners’
Knowledge of Web Technologies
and Their Uses
Chief Holmes had just suffered through an hourlong presentation on the ever-worsening cyber threat. He knew about hijacked
and defaced websites, but that was just the very tip of the iceberg.
Stolen case investigation records. Names, photos, and addresses of
undercover officers posted online. Overseas crime syndicates holding
departments’ entire computer systems for ransom. That part was
terrifying, but at least it was understandable. Then came half an
hour of numbing acronym soup about dozens of places to report
attacks, dozens of development tools, government standards, references that presumably were great if you were a computer security
expert, and something about improved tags that would help protect
data and intelligent agents that would help figure out who was a
real user and who was a hacker. When question time finally came
about, he raised his hand.
“Excuse me, but like most of us, all I know about security is
to install antivirus and not use the return key for my password.
You’ve done a great job telling us about the threat, but for those of
us who aren’t security experts, what are we supposed to do? How do
we get started?”
“Well . . . um . . . maybe you could go to . . . um . . . ?”
———
The highest-ranking machine-to-user need—and the
highest-ranking information-sharing need overall—was to
develop real-time language interpretation services.
14
This theme comprises methods to help practitioners learn
more about technologies, use them properly and effectively,
and acquire them. Improving practitioners’ knowledge had the
second-highest number of top-tier needs (five)
• educational materials on key web technologies (ranked
second overall)
• cost-benefit tools for technology acquisition
• policies and procedures for emerging IoT and Semantic
Web technologies
• procurement checklist for courts IT (ranked second for courts)
• policies and procedures for unmanned, automated vehicles
(top-ranked need for law enforcement).
Education on Web Technologies, in General
The second-ranked need overall was to provide the criminal
justice community with educational materials on the core
Web 3.0+ technologies. This need is broadly in keeping with
the findings from prior research into improving practitioners’
knowledge of key technologies in general. Panelists discussed
providing education and training, model policies and procedures, and/or acquisition checklists for
• Semantic Web technologies (and related developments,
such as document databases) that specifically support finding criminal justice information (such as criminal histories), getting access to information, and translating and
using information. An example is the discussion in the previous section of how to use semantic tags combined with
document databases to share criminal justice histories.
• IoT sensors and actuators, especially those related to top
needs for monitoring officer health and officer safety, as
well as maintaining house arrest conditions (as described
under Theme 4, below)
• entity analytics, which are tools that can take plain text
and parse them into structured data. Key examples include
tools that can pull names, addresses, phone numbers,
license plate tags, and email addresses out of plain text, as
well as generate relationships between them (such as details
about people who live at the same address).
• video conferencing/tele-education technologies and services,
to include both the tools themselves and the infrastructure
and information access policies needed to support them
• civil rights, privacy rights, and cybersecurity technology
issues and solutions (see Theme 5, below)
• emerging real-time language systems, as discussed above.
The principal opportunity in web technologies education is
that a great deal is known about these technologies, especially
in the commercial and academic sectors. However, there are
two principal barriers. The first is that these technologies are
largely new and unfamiliar to criminal justice, and it will take
time and effort to introduce them.
The second is that technology education material is very
widely distributed, with the Department of Justice (DoJ),
practitioner associations, and some commercial associations all
maintaining a wide range of web portals on criminal justice
technology with varying degrees of awareness by the criminal
justice communities of practice (see Gordon et al., 2012, for
a larger discussion of this issue, focusing on the DoJ side).
Even within portals, material is commonly presented as lists
or databases of disjointed documents. At the 2015 Workshop
on Information Sharing and Safeguarding, for example, it
was noted that two key portals on information sharing tools,
standards, and methods—the Standards Coordinating Council
(SCC) portal (2015) and the Project Interoperability (undated)
portals—are new and can thus be reasonably described as
“grab-bags of stuff,” with sponsors planning to integrate the
materials and their presentation over time. These and other
portals work for technical experts seeking a specific reference or
two but do not work well for those seeking to develop technical
expertise in the area or, more broadly, for agencies looking to
acquire new technologies without much prior background.
Specific Needs Related to Technology Knowledge
On the specific needs side, the procurement checklist for courts’
IT scored highly; it has been mentioned that the National
Center for State Courts and others are preparing procurement
guidance. That guidance may end up largely meeting this need.
The top law enforcement priority was developing policies and
procedures for self-driving unmanned and automated vehicles. In
addition to being the top-ranked need for law enforcement, it
was the most “futuristic” of the needs considered. However, this
area is developing rapidly; for example, California in 2014 issued
to Audi the state’s first permit allowing autonomous vehicles on
public roads (Franzen, 2014). Panelists noted that law enforcement work in this area should coordinate closely with major
existing policy efforts under way by the National Highway
Traffic Safety Administration (NHTSA), state departments of
transportation (especially the California Department of Motor
15
The breakout need from the workshop was to provide
video links to corrections facilities so that community
corrections officers could meet with inmates prior to their
release.
Vehicles [DMV], given the amount of self-driving car activity
centered there; see Madrigal, 2014), top self-driving car developers, and others. Panelists also advised prudence: The lowest-ranking need overall was to develop an interface for officers to directly
take control of unmanned vehicles.
Theme 3: Improve Infrastructure
“And that about does it for me. Mr. Smith. Big thing is, our
first phone call meeting at 2 p.m. next Thursday. You’re going to
call in from the warehouse lounge, right?”
“Yes.”
Another voice and face chimed in on the screen. “And I’ ll
expect you bright and early for your first day on the job on Thursday at 8:30 a.m., right? You know how to get here? If not, just call
and we’ ll answer any questions you have.”
“Yes, I won’t have any problem.”
A third person spoke. “And on Friday, you need to come to
your initial counseling appointment at the Center after work. I
know you know where that is.”
“Yes, Dr. Nassar, looking forward to it.”
“Okay, sounds like we’re done here. Good luck and congratulations on getting early parole. Don’t blow it. See you next week.”
With that, the screen on Mr. Smith’s tablet went blank. He
leaned back on his cell cot, and pressed another tablet button to
bring up the day’s sports scores.
———
The common element in this theme is that leveraging web technologies in criminal justice requires the networking infrastructure to support them. This theme includes only three top-tier
needs, but these were some of the highest-ranking needs. They
include
• video links to correctional facilities (top need overall, for
corrections and for courts)
• high-speed Internet connections for courts
• virtual courtrooms.
We talk below about each of these needs individually.
The breakout need from the workshop was to provide
video links to correctional facilities so that community corrections officers (and other stakeholders, such as service providers)
could meet with inmates prior to their release. We are aware of
existing videoconferencing links and supporting technologies
(secure tablets, for instance) to support remote education for
prisoners; it may be possible to leverage some these systems for
meetings with community corrections officers and other service
providers. Notably, in 2014, the White House Office of Science
and Technology Policy held a workshop on using technology
to improve inmate re-entry. One of the major discussion topics
at the workshop was the use of remote learning technologies
(including video teleconferencing) to provide educational services to inmates; the technologies and infrastructure discussed
there might be expanded to include teleconferences with external probation and parole officers and service providers.
The other two needs concerned high-speed Internet buildouts to courts. The first involved providing high-speed connections to existing courthouses in rural areas; the second (slightly
lower-ranked) involved creating mobile communications
packages to courtrooms that can be set up in other locations
as needed. One panelist noted that the virtual courtrooms
could be lower-ranked if high-speed Internet were provided
to existing rural courthouses, because much of the need for
virtual courtrooms and their associated expenses would disappear. These needs were primarily about obtaining funding for
technology improvements; not much R&D is needed here. The
principal barrier is obtaining funding.
Theme 4: Exploring the Use of Emerging
IoT Sensors in Criminal Justice
Officer Jackson thought she smelled something—kind of
like rancid cigarette smoke—but didn’t think much of it as she
searched the suspected drug house for stacks of small bills. She
reached out to open a door that led down a hall to the kitchen.
16
Three of the four needs
concerned furthering
officer health and safety.
The door felt warm, but not enough to really ring any alarm bells
in her head. Suddenly, an alarm sounded in her ear, and a voice
shouted over her earpiece, “Don’t open that door! The house’s alarm
system says there’s a fire in the kitchen!”
———
Of the three major Web 3.0+ technology areas (IoT, intelligent
agents, and Semantic Web), IoT proved to be the one with its
own cluster of needs. In comparison, Semantic Web technologies were subsumed into information sharing, and needs related
to intelligent agents (search agents and decision agents) tended
not to rank highly.
This theme includes four top-tier needs and has the fourthgreatest sum of overall expected value across its needs. Needs in
this cluster constituted 8.7 percent of all needs’ overall expected
value scores. Top needs are
• Internet-enabled biomedical sensors for officers. A great
deal of academic and commercial development has gone on
this area; Steele and Clarke (2013) provide a reference.
• improving the tracking of officers within buildings. A great
deal of academic and commercial development has gone
on in this area, too; Schutzberg (2013) provides a quick
reference.
• IoT-enabled models for house arrest
• identify officers in close proximity. The panelists noted that
the human factors issues would require special attention—
for example, one does not want loud buzzing to identify
that an undercover agent is present.
Of interest, three of the four needs concerned furthering
officer health and safety, which were similar to clusters of IT
needs for law enforcement. Thus, protecting the health and
safety of officers appears to be the principal topic of interest
in this area. The final need involved the use of further sensors
to supplement body-worn Global Positioning System (GPS)
devices to locate and track offenders who are under community
corrections supervision.
The use of IoT sensors in criminal justice involves two
principal barriers. The first is the comparative newness of the
technologies, with the exception of early body-worn GPS tracking devices. A great deal would need to be done to introduce
and experiment with the technologies, and we reiterate that the
challenge includes not just the devices themselves but also the
large IT back-end needed to get data from the devices and use
those data effectively.
The second comprises the risks to civil rights, privacy
rights, and cybersecurity raised by the widespread use of
deployable sensors. As noted below, IoT sensors received
focused attention for raising these risks, and special attention
will need to be paid to mitigating those risks.
Theme 5: Civil Rights, Privacy Rights, and
Cybersecurity Concerns
The hacker grinned at his monitor. A few more commands
and all of the supposedly secret names, addresses, and photos of
the city’s undercover officers and informants would be his. The
initial plan was to dump all that information online. But then he
wondered how much he could get by selling it to the local crime
syndicates. Very tempting.
At last he was in. He eagerly downloaded the file and laughed
as he brought it up. What? Big Bird? Bozo the Clown? Was this
someone’s idea of a joke? If so, he’ d show them what a joke was
when he made them pay—after a good night’s sleep.
In the middle of the night, he awoke to what he wished was
simply a nightmare: a group of officers who had been given his
identity and location by the city’s cybercrime unit.
———
Two overarching concerns were frequently discussed at the
workshop, although they did not make their way into many
needs. The first involved cybersecurity concerns arising from
the new technologies, and the second involved privacy rights
and civil rights.
Figure 2 shows the proportions of needs that panelists
flagged as having potential civil rights, privacy rights, or cybersecurity risks. Twenty-four needs (53 percent) were flagged
for potential civil rights or privacy rights implications; nine
(20 percent) were flagged for potential security implications.
From a civil and privacy rights perspective, search agents
were called out as potentially problematic due to their analyzing large amounts of public data (public surveillance cameras
and social media feeds, for example) and generating large num-
17
Figure 2. Civil and Privacy Rights Risks and Cybersecurity Risks
Civil and privacy rights risks
Other
17%
Search agents
(event detection)
25%
IoT sensors
(wearable/
embedded)
21%
Policies,
procedures,
and guidance
12%
Decision
agents
8%
Standards
17%
Cybersecurity risks
Other
22%
Standards
45%
IoT sensors
(wearable/
embedded)
22%
Decision
agents
11%
RAND RR928-2
bers of false positives (such as subjects being incorrectly flagged
as likely engaged in criminal behavior).
Decision agents, such as those related to recommending
bail/bond or sentencing judgments, were seen as potentially
raising civil rights concerns for computing decisions based on
inadequate or biased data.
IoT sensors were seen as raising potential privacy issues
from generating substantial data on persons either wearing sensors or those who are nearby. For example, one panelist noted,
“We may need to monitor someone under community supervision, but we should not end up monitoring the others at the
group counseling session with them.”
From a security perspective, IoT sensors and other devices
were seen as potentially vulnerable. By definition, they transmit
and receive a great deal of data, providing access for hackers, and
their small, lightweight, and inexpensive nature implies that the
devices will not have a great deal of computing power to run
security software. Beyond cybersecurity, operational security for
sensor data was raised as well. For example, the top-tier need for
sensors that can detect when other officers are nearby was seen as
being a potentially critical tool for crisis response and operational
deconfliction but also as a threat if criminals could recognize an
undercover agent’s sensor during an operation.
Standards were not seen as posing inherent risks to civil
rights, privacy rights, or security. Rather, the concern was that
because standards are inherently about enabling the sharing of
sensitive criminal justice information, strong security measures
and strong civil and privacy rights protections should be built
into the standards to avoid opening vulnerabilities. Similarly,
policies and procedures for using information resulting from
Web 3.0+ technology applications were seen as needing strong
protections for civil rights, privacy rights, and cybersecurity to
defend against misuse of sensitive information.
Civil rights, privacy rights, and cybersecurity protections
are major pain points across criminal justice technologies in
general. They are repeatedly described at criminal justice workshops and conferences as areas in which agencies lack expertise.
The White House Cybersecurity Coordinator, for example,
described information security in general (not just for criminal
justice) as widely seen as too hard and too expensive, and as
something that still has to be added on by experts rather than
being designed in and turned on in hardware and software
systems by default (Daniel, 2015).
Comprehensive List of Key Criminal Justice
Needs
Table 3 presents the needs from the Web 3.0+ Technologies
Workshop by their technology category. Within tables, needs
are presented in order of their overall expected value score,
with highest score first. We also show what tier the need is in
with respect to overall expected value. The last two columns
show whether panelists flagged the need as having civil rights,
privacy rights, or security issues.
18
Table 3. Needs from the Web 3.0+ Technologies Workshop
Legend
Community of Practice key:
Law enforcement
Tier key:
1: Top priority
Privacy Rights/Civil Issue and
Security key:
✓ Concern identified by panelists
Community of
Practice in Need
Courts
Corrections
2: Medium priority
Description
3: Low priority
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 1: Decision Agents
Digital evidence chain of custody
Need intelligent agents to help protect the chain of custody as
digital evidence receives metadata markups for semantic search and
analysis
1
Skills assessment
Need to develop tools to help incident commanders/managers
quickly identify or assemble teams with specific skills (such as
language, hazardous materials, or hostage negotiation skills)
2
Managing court workload
Need cognitive/artificial intelligence systems that use a range of
data, including real-time data feeds, to inform day-to-day workloads
of criminal justice staff ranging from law enforcement officers to
judges
2
Interviews and report writing
Need intelligent agents to assist with field interviews and writing
incident reports, along with the controlled vocabularies required
to develop them. The agents should make it easy to enter data,
structure the data as they are entered to support later queries, and
help enforce completeness (for example, “This is a burglary and
you haven’t entered a mode of entry. If you know the mode of entry,
enter it now.”)
2
Digital evidence gathering
Need intelligent agents that can guide patrol officers to identify and
recover digital evidence
2
Offender risk assessment
Need new risk assessment tools to inform court decisions and
evaluate existing ones. These include tools informing pretrial bail
setting and post-trial sentencing. Should include defining metrics and
best practices for risk assessment tools
2
✓
Support for specialized courts
Need intelligent agents as “expert systems” to support decisionmaking for specific courts (drug courts, veteran courts, etc.)
2
✓
✓
19
Table 3—Continued
Community of
Practice in Need
Description
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 2: Internet-of-Things Sensors
Biomedical sensors for officers
Need to develop and/or assess wearable biomedical sensors to
monitor officer health and safety. Sensors should monitor stress
levels, fatigue, and injuries. Data could be used to dynamically
shorten work shifts if fatigue levels are excessive
1
Tracking officers inside buildings
Need to study using IoT sensors to better track officers inside
buildings. Sensors should be both officer-worn sensors and within
the building (for example, proximity sensors). Tracking would help
improve officer safety and could be used for opening doors and
other applications
1
Monitoring alcohol/drug use
Need to develop and/or assess wearable sensors for monitoring
offender use of alcohol or drugs as well as procedures for
employing them effectively. Research should include authentication
(ensuring the right individual is wearing the device) and integration
of data into dashboards
2
✓
Monitoring house arrest
Need to experiment with using IoT sensors to better enforce house
arrest and noninstitutionalized corrections options
2
✓
Identify nearby officers
Need to develop and/or assess wearable proximity sensors (such as
automated virtual badges) to improve officers’ situational awareness
of each others’ locations. Could be used for task forces; high-density,
poor visibility, operational deconfliction; and undercover work
2
✓
Emergency help for at-risk civilians
Need sensors to detect when someone needs emergency help at
home (for people with disabilities or disease, or individuals at risk of
domestic violence)
3
✓
Part 3: Miscellaneous Technologies
Video links to correctional facilities
Need to develop and/or assess video teleconferencing tools to
allow incarcerated offenders to establish relationships and build
rapport with community corrections officers before release
1
Real-time translation
Need the capability to translate languages in real time. If not standalone, the capability should include connections so that remote
officers and courts can access translation services
1
✓
✓
✓
20
Table 3—Continued
Community of
Practice in Need
Description
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 3: Miscellaneous Technologies (continued)
High-speed Internet
Need to fund high-speed Internet for courts to enable web
technologies and other tools
1
Facial recognition search across databases
Need to develop a system to search across multiple mug shot
databases and other photo databases at the federal, state, and
local levels
1
Information overload research
Need research on the impact of information overload on law
enforcement, courts, and corrections personnel, as well as the
causes of distractions and potential solutions
1
Situational awareness tools
Need to develop technologies to provide situational awareness
displays (annotated maps, “data mashups,” and customized alerts)
dynamically tailored to individual officers both in the field and at
headquarters. Should automatically generate alerts when people
of interest (such as parolees) have made contact with police.
Technologies will include developing better data models
1
Extracting data from reports
Need tools to extract structured data (names, addresses, phone
numbers, vehicle plates, etc.) from narrative descriptions about
incidents and people (such as incident descriptions in the FBI’s
National Data Exchange [N-DEx] program, a repository of criminal
justice records)
2
Virtual courtrooms
Need to develop and/or assess portable courtroom video
teleconferencing “kits” that judges can set up nearly anywhere (such
as a public library) to alleviate travel burdens in small and rural
settings
2
Court case management
Need to further develop and evaluate integrated court case
management systems (CMS) that provide case information across
a large set of courtrooms (such as statewide systems). Development
might be done through entirely new builds, rehabbing existing
CMS, or adopting and integrating law enforcement and corrections
records systems. Evaluation should include defining metrics and best
practices for emerging CMS
2
✓
Code-breaking tools
Need tools to decipher codes used by gang members and other
offenders to conceal their communications. One approach might be
to use natural language models
2
✓
Controlling automated vehicles
Need methods for control and manipulation of automated vehicles
(when warranted). For example, law enforcement may want to direct
a parked vehicle to move
3
✓
✓
✓
✓
21
Table 3—Continued
Community of
Practice in Need
Description
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 4: Policies, Procedures, and Guidance
Technology training
Need educational materials for key web technologies, including
Internet-enabled sensors and actuators, entity analytics, data
life-cycle management, video conferencing, translation, and
tele-education. Materials might include training, trade show
presentations, and workshops
1
Cost-benefit tool for technology acquisition
Need cost-benefit research to assess the return on investment on
technology-related acquisitions/programs. The research should
produce not only general guidance on which investments are
valuable but also tools to assess specific investment options, as
both are needed to help determine and communicate the value of
possible technology investments
1
Policies and procedures for emerging technologies
Need policies, tactics, rules of engagement, and best practices for
responding to emerging technologies such as IoT and Semantic Web
technologies. Should address policies around the collection, storage,
and transfer of data from IoT-enabled objects (cars, appliances, etc.)
1
Procurement checklist for courts’ IT
Need to develop procurement checklists for courts’ IT, including:
(1) discouraging vendor lock-in resulting from acquiring proprietary
systems, (2) advance documentation of data formats and software
interfaces, and (3) benchmark and performance assessments to help
ensure that systems perform as expected
1
Aerial drones and driverless vehicles
Need policies, tactics, rules of engagement, and best practices for
responding to emerging technologies such as aerial and terrestrial
automated and unmanned vehicles. This includes methods to extract
data from IoT-enabled vehicles for improved accident investigation
and reporting, educating practitioners on the capabilities of new
technologies, friend or foe identification, locating the owner/
operator, and techniques for safe neutralization.
2
✓
Security of IoT sensors
Need to examine security implications and countermeasures against
the use of IoT sensors by criminals and the public to collect and
publicize information on criminal justice operations (for example,
tracking officer locations, identifying undercover officers)
2
✓
✓
22
Table 3—Continued
Community of
Practice in Need
Description
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 5: Search Agents
Monitoring criminals’ associations
Need intelligent agents to detect criminal activity by monitoring
offenders’ activities and social connections (such as social network
analysis)
2
✓
Detecting patterns of criminal activity
Need intelligent agents to search online data for activity patterns
worthy of additional attention (for example, human trafficking and
money laundering)
2
✓
Video analytics
Need to improve the accuracy of existing video analytic tools,
to include integrating the tools with other types of sensor data to
improve detection of serious events
2
✓
Semantic search assessment
Need to assess existing digital evidence toolkits to determine
whether semantic search tools could substantially speed up
searching and analyzing digital evidence
2
Analytics for social media and community feedback
Need automated analytics of social media and community board
postings to assess public feedback and ambient community safety
(“the community as an intelligent sensor”)
2
Sensor-detected crime
Need to develop and/or integrate data from audio, video, and
other sensors carried by civil servants, civil fleet vehicles, and/or
volunteers to identify incidents worthy of additional investigation
(such as gunshots)
3
✓
Sharing offender tracking data
Need research on detecting and sharing events of interest in
offender tracking data. Examples include meeting with treatment
providers, associating with other offenders, and being near a
probation or law enforcement officer. Research should include
methods for improving tracking accuracy
3
✓
Semantic analysis of inmates’ phone calls
Need automated audio analytics to assist with monitoring inmate
communications
3
✓
23
Table 3—Continued
Community of
Practice in Need
Description
Tier
Privacy Rights/ Security
Civil Issue?
Issue?
Part 6: Standards
Virtual criminal record catalog
Need a common criminal history record, a common catalog for
storing the records, policies governing access to the records, and
business processes for ensuring the records are up to date, correct,
and consistent.
1
Technological infrastructure
Need technologies to support data exchange across the criminal
justice enterprise. These include agreed-to data formats and
Application Protocol Interfaces (APIs), a set of routines and tools for
building software applications; authentication/identity management;
access control mechanisms; cloud; and accessible-anywhere tools
1
Tools to improve data quality and integrity
Need research on methods to enhance data quality and integrity,
including provenance markings (PROV standard; integrated into
NIEM), collection criteria, privacy attributes and other self-enforcing
data-usage policies, use of intelligent agents for continuous data
validation, and metadata matching
✓
✓
1
✓
✓
Information sharing
Need methods to better share information about offenders between
corrections agencies and third parties (e.g., treatment providers and
other stakeholders)
1
✓
✓
Common technology vocabulary and use cases
Need a core set of use cases that describes where and how
criminal justice activities should be supported with IT, along with
ontologies providing the “grammar” for doing so. (Here, an ontology
is a formal vocabulary that describes the types, properties, and
interrelationships of criminal justice concepts)
2
Emergency data sharing
Need common data models and data access mechanisms that will
support integrating data feeds from multiple sources (including IoT
sensors from the public and private sectors) and providing the data
to multiple types of stakeholders (Red Cross, etc.) in response to
unanticipated needs. Should begin by considering what semantic
extensions NIEM needs
2
✓
✓
Metadata standards for video to support semantic search
Need to determine standards (“ideals”) for labeling video with
metadata tags to support semantic search and analysis
2
CO N C LUSI O NS: SET TI N G TH E
TEC H N O LO GY AGE N DA
In this concluding section, we present ways to take action on
the themes and needs that can help criminal justice leverage the
new web technologies.
Leveraging the New Web Technologies to
Improve Information Sharing
Improving information sharing in general. As noted, both
the major opportunity and major challenge is the large number of existing efforts by a variety of organizations to improve
criminal justice information. The SCC has been established to
24
As with information sharing, a number of ongoing
efforts are disseminating knowledge about technology,
technology acquisition, and technology policy education
and guidance efforts.
integrate the work of standards-making organizations. Indeed,
one purpose of the council is to “identify high-priority standards activities that can be coordinated across standards development organizations (SDOs) for greater return on resources”
(Information Sharing Environment, undated). The next step,
then, would be to request that the council and its members
explore how semantic technologies, as well as intelligent agents
that exploit semantic tags, might be leveraged to make it much
easier for developers to make criminal justice information
searchable and available, while also ensuring its safeguarding
(to include data quality, integrity, and security).
The top-ranked need, developing a virtual criminal history, would be a good place to start. In addition to being a
highly ranked need in its own right, criminal history information frequently is used in many applications across the
criminal justice enterprise, ranging from an officer in the field
querying for data about a stopped driver, to a judge looking
up a defendant’s prior criminal history prior to a bail or bond
decision, to institutional and community corrections officers
using criminal history background to make risk prioritization
decisions.
Better access to mug shots and other facial data. The
next step here is to work with the FBI to develop standards
and common policies for sharing mug-shot photos in response
to queries, consistent with the FBI’s larger Next Generation
Identification system. A key aspect will be policies and procedures to protect against abuses, as well as to mitigate the risks
of false positive identifications. The new Trustmarks (GTRI
NSTIC Trustmark Pilot, undated) and other security tagging
technology might be leveraged here to expedite setting up
memoranda of understanding for sharing mug shots and other
facial data.
Tailored situational awareness displays and reducing information overload. The next step for these needs is to
explore folding web semantic tagging and intelligent agents
that populate displays into existing R&D. As noted, semantic
tags and intelligent agents might lead to better ways to find
and display what is most relevant in various situations while
hiding what is irrelevant. One way is to have a subsidiary effort,
under the larger proposed SCC initiative, explore using tagged
data and agents to populate situational awareness displays (also
known as dashboards or common operational pictures) at multiple
levels.
Developing real-time language interpretation services.
The next step here is to experiment with the new commercially
available services, such as Microsoft’s new English-to-Spanish
translation service (Warren, 2014).
Improving Practitioners’ Knowledge of New Web
Technologies and Their Uses
Education on web technologies in general. As with information sharing, a number of ongoing efforts are disseminating
knowledge about technology, technology acquisition, and technology policy education and guidance efforts. The natural next
step here is to determine which organizations (or individuals)
might be called upon to develop educational material, along with
model policies and checklists. As with information sharing technologies, it would be useful to start with the SCC to assess who
is coordinating policy and educational efforts for these technologies. Specific technologies to focus education efforts on include
• Semantic Web technologies (and related developments,
such as document databases) that specifically support finding criminal justice information (such as criminal histories), getting access to information, and translating and
using information
• IoT sensor systems related to monitoring officer health and
officer safety
• IoT sensor systems that could be used to maintain house
arrest and other community supervision conditions
25
• entity analytics, which are tools that can take plain text
and parse it into structured (tagged) data. This includes
tools that can pull names, addresses, phone numbers,
license plate tags, and email addresses out of plain text, as
well as generate relationships between them (for example,
details about people who live at the same address).
• video conferencing/tele-education technologies and services,
to include both the tools themselves and the infrastructure
and information access policies needed to support them
• civil rights, privacy rights, and cybersecurity technology
issues and solutions (discussed further below)
• emerging real-time language systems (as discussed above).
The more general next step is to begin converting key
technology portals from lists of disjointed materials to learning
portals, which provide both technical developers and agencies
seeking to acquire technologies with a logical series of instructional materials to get the knowledge they need. As discussed in
the previous section, this process is just starting in the criminal
justice information sharing area.
Policies and procedures for self-driving cars. The topranked need in this category—policies and procedures for
self-driving cars—will need close coordination with the existing policy efforts under way by NHTSA, state departments of
transportation, and others. The next step here is to designate a
group from the law enforcement community to take the lead
in partnering with the key stakeholders (NHTSA, California
DMV, industry) to (1) develop law enforcement requirements
for self-driving vehicles, (2) ensure law enforcement requirements are included both in regulations and in technical design
specifications, and (3) develop policies and procedures for law
enforcement personnel in dealing with self-driving cars.
Improving Infrastructure
The breakout need of the workshop was video teleconferencing
with prisoners soon to be released. Related technologies supporting remote education for prisoners are sufficiently advanced
to support developing and conducting field experiments on
video teleconferencing links. The next step here is to designate a
group from the criminal justice community to
• identify use cases and requirements for video teleconference (VTC) technology between inmates and officers and
external service providers
• develop standard policies and procedures for using VTC,
leveraging the information sharing safeguarding policies
discussed previously
• identify and contact technology providers and funders
about what might be possible to provide at what cost, and
disseminate the results to the larger criminal justice community, including practitioners, developers, and funders.
The other two top needs in this area are largely technologically solved. The next step is to designate a group to identify a
standard set of requirements for high-speed Internet to rural
courthouses, along with standard estimated costs, which agencies then could use agencies for both planning purposes and
to seek funding. The group should also pursue novel business
models and support to make Internet links more affordable in
rural areas.
Using IoT Sensors in Criminal Justice
Two directions of applied development and experimentation
are recommended. The first is experimenting with health and
safety sensor feeds—both wearable sensors and those embedded
in buildings. The second is experimenting with Internetconnected sensors (and supporting Semantic Web and intelligent agents for sharing and analyzing the data feeds) to support
the location and tracking of offenders under community corrections supervision. Technologically, this means experimenting in two directions. The first is on tracking systems, both
wearable and contained in infrastructure, that can leverage the
substantial technologies already developed. The second is on
biomedical sensors, again experimenting with the substantial
technologies already developed.
The breakout need of the workshop was video
teleconferencing with prisoners soon to be released.
26
Addressing Civil Rights, Privacy Rights, and
Cybersecurity Concerns
Finally, given the high concern about civil rights, privacy rights,
and cybersecurity risks, there are next steps to be taken.Most
important is ensuring that civil rights, privacy rights, and cybersecurity provisions are incorporated into emerging informationsharing standards, policies, and procedures that deal with the
emerging web technologies. At their core, these are strategic,
national-level issues that need to be inculcated throughout
criminal justice development efforts. Developing a full civil
rights, privacy rights, and cybersecurity strategy is well beyond
the scope of this report. However, there are a few initial points:
• Note White House guidance that cybersecurity (and IT
privacy rights and civil rights protections, relatedly) needs
to move from being too difficult and too expensive for agencies (and other organizations) to add to their systems—they
need to be built in and turned on by default (Daniel, 2015).
• The International Association of Chiefs of Police has created a general Technology Policy Framework that describes
common provisions that should be part of any policies
and operating procedures for new technology acquisition
and use, including civil rights, privacy rights, and security
protections (International Association of Chiefs of Police,
2014). This framework can be built upon for future technology development, experimentation, and fielding efforts.
Regarding intelligent agents and IoT sensors, two web
technologies that were repeatedly flagged as raising special civil
rights, privacy rights, and cybersecurity concerns:
• For intelligent agents supporting criminal justice decisionmaking about people (who should be the focus of law
enforcement or community corrections scrutiny, bail/bond
decisions, sentencing decisions, etc.), we recommend a study
on how to ensure the quality of data used to make the decision, as well as how decisionmakers should use the agents’
recommendations given realistic uncertainties and potential
biases.
• For IoT sensors and search agents retrieving data across
both sensor feeds and data stores, the next step is a study
to develop common attributes for criminal justice policies,
procedures, and required protective technologies for both
sensor feed access control and cybersecurity.
For both studies, criminal justice technology experts and experts
on civil rights, privacy rights, and information security should
be participants.
27
Notes
1
A precision immobilization technique (PIT) is a driving maneuver designed to cause a suspect’s vehicle to spin out and stop.
2
IJIS formerly stood for “Integrated Justice Information Systems.” It now simply stands for IJIS.
This has to do with many IEPDs being from small information-sharing efforts, in which a few agencies document the specific standards they
are using to share information with each other.
3
28
Bibliography
Ahmed, Alijah, “K-Means Clustering Calculator,” SciStat Calc, January 24, 2014. As of June 12, 2015:
http://scistatcalc.blogspot.com/2014/01/k-means-clustering-calculator.html
Bir, C., J. Cecconi, A. Dennis, M. McMullen, and C. Sloane, Behind the Badge: Management Guidelines for Impacts to Body Armor,
National Institute of Justice, Award Number 2004‐IJ‐CX‐K040, 2011. As of June 12, 2015:
https://www.ncjrs.gov/pdffiles1/nij/grants/233645.pdf
Braga, Anthony A., Andrew V. Papachristos, and David M. Hureau, “The Effects of Hot Spots Policing on Crime: An Updated Systematic
Review and Meta-Analysis,” Justice Quarterly, iFirst, 2012, pp. 1–31.
Contchart Software, “Outlier Tests for the Normal Distribution,” ControlFreak: Control Charts for Individual Data Points, 2014. As of
June 12, 2015:
http://contchart.com/outliers.aspx
Daniel, J. Michael, presentation to the 2015 Workshop on Information Sharing and Safeguarding, Reston, Va., March 26, 2015.
Foaf Project, homepage, undated. As of June 12, 2015:
http://www.foaf-project.org/
Franzen, Carl, “Google’s Self-Driving Cars and Others Get Permits to Drive in California,” The Verge, September 22, 2014. As of July 8, 2015:
http://www.theverge.com/2014/9/22/6828161/california-permits-self-driving-cars-google-audi-mercedes-benz
Gordon, John, IV, Brett Andrew Wallace, Daniel Tremblay, and John S. Hollywood, Keeping Law Enforcement Connected: Information
Technology Needs from State and Local Agencies, Santa Monica, Calif.: RAND Corporation, TR-1165-NIJ, 2012. As of July 7, 2015:
http://www.rand.org/pubs/technical_reports/TR1165.html
GTRI NSTIC Trustmark Pilot, homepage, undated. As of June 12, 2015:
https://trustmark.gtri.gatech.edu/
Heflin, Jeff, Raphael Volz, and Jonathan Dale, eds., “Requirements for a Web Ontology Language” (working draft), World Wide Web
Consortium, March 7, 2002. As of June 12, 2015:
http://www.w3.org/TR/2002/WD-webont-req-20020307/
Hollywood, John S., John E. Boon, Jr., Richard Silberglitt, Brian G. Chow, and Brian A. Jackson, High-Priority Information Technology Needs for
Law Enforcement, Santa Monica, Calif.: RAND Corporation, RR-737-NIJ, 2015. As of June 12, 2015:
http://www.rand.org/pubs/research_reports/RR737.html
IEPD Clearinghouse, homepage, undated. As of July 7, 2015:
http://iepd.custhelp.com/
Information Sharing Environment, Standards Coordinating Council, undated. As of June 12, 2015:
http://www.ise.gov/standards-coordinating-council
International Association of Chiefs of Police, Law Enforcement Priorities for Public Safety: Identifying Critical Technology Needs, Alexandria, Va.,
2005.
———, IACP Technology Policy Framework, January 2014. As of June 12, 2015:
http://www.theiacp.org/Portals/0/documents/pdfs/IACP%20Technology%20Policy%20Framework%20January%202014%20Final.pdf
Jackson, Brian A., Joe Russo, John S. Hollywood, Dulani Woods, Richard Silberglitt, George B. Drake, John S. Shaffer, Mikhail Zaydman, and
Brian G. Chow, Fostering Innovation in Community and Institutional Corrections: Identifying High Priority Technology and Related Needs for the
United States Corrections Sector, Santa Monica, Calif.: RAND Corporation RR-820-NIJ, 2015. As of June 12, 2015:
http://www.rand.org/pubs/research_reports/RR820.html
Kelly, Kevin, “The Next 5,000 Days of the Web,” TED.com, December 2007. As of June 12, 2015:
http://www.ted.com/talks/kevin_kelly_on_the_next_5_000_days_of_the_web
29
Koper, Christopher S., Bruce G. Taylor, and Bruce E. Kubu, Law Enforcement Technology Needs Assessment: Future Technologies to Address the
Operational Needs of Law Enforcement, Washington, D.C.: Police Executive Research Forum and Lockheed Martin Corporation, 2009. As of
July 8, 2015:
http://www.policeforum.org/assets/docs/Free_Online_Documents/Technology/law%20enforcement%20technology%20needs%20
assessment%202009.pdf
Law Enforcement Information Technology Standards Council, Standard Functional Specifications for Law Enforcement Records Management
Systems Version II, Washington, D.C.: Bureau of Justice Assistance, 2008.
Lerman, Julie, “What the Heck Are Document Databases?” MSDN Magazine, November 2011. As of June 12, 2015:
https://msdn.microsoft.com/en-us/magazine/hh547103.aspx
Madrigal, Alexis C., “Meet the Regulators Trying to Make Sure Self-Driving Cars Are Safe,” The Atlantic, May 22, 2014. As of July 8, 2015:
http://www.theatlantic.com/technology/archive/2014/05/meet-the-regulators-tasked-with-making-sure-self-driving-cars-are-safe/371352/
National Institute of Standards and Technology, “Grubbs’ Test for Outliers,” e-Handbook of Statistical Methods, 2013. As of June 12, 2015:
http://www.itl.nist.gov/div898/handbook/eda/section3/eda35h1.htm
Page, Lawrence, Sergey Brin, Rajeev Motwani, and Terry Winograd, The PageRank Citation Ranking: Bringing Order to the Web (working paper),
Stanford InfoLab Publication Server, January 29, 1998. As of June 12, 2015:
http://ilpubs.stanford.edu:8090/422/
Project Interoperability, homepage, undated. As of June 12, 2015:
http://project-interoperability.github.io/
Schutzberg, Adena, “Ten Things You Need to Know About Indoor Positioning,” Directions Magazine, May 6, 2013. As of June 12, 2015:
http://www.directionsmag.com/entry/10-things-you-need-to-know-about-indoor-positioning/324602
Silberglitt, Richard, Brian G. Chow, John S. Hollywood, Dulani Woods, Mikhail Zaydman, and Brian A. Jackson, Visions of Law Enforcement
Technology in the Period 2024–2034: Report of the Law Enforcement Futuring Workshop, Santa Monica, Calif.: RAND Corporation, RR-908-NIJ,
2015. As of June 12, 2015:
http://www.rand.org/pubs/research_reports/RR908.html
Smith, Rick, “Why Is Public Safety Behind the Tech Curve?” Police, April 9, 2013. As of June 12, 2015:
http://www.policemag.com/blog/technology/story/2013/04/why-is-public-safety-behind-the-tech-curve.aspx
Standards Coordinating Council, SCC Member Organizations, undated. As of June 12, 2015:
http://www.standardscoordination.org/scc-member-organizations
Steele, Robert, and Andrew Clarke, “The Internet of Things and Next-Generation Public Health Information Systems,” Communications and
Networks, Vol. 5, No. 3B, 2013, pp. 4–9.
U.S. Department of Justice, NIEM IEPD Clearinghouse, 2014. As of June 12, 2015:
https://it.ojp.gov/framesets/iepd-clearinghouse-noClose.htm
Wadwha, Tarun, “Lessons from Crowdsourcing the Boston Bombing Investigation,” Forbes, April 22, 2013. As of June 12, 2015:
http://www.forbes.com/sites/tarunwadhwa/2013/04/22/lessons-from-crowdsourcing-the-boston-marathon-bombings-investigation/
Warren, Tom, “Skype’s Newest App Will Translate Your Speech in Real Time,” The Verge, December 15, 2014. As of June 12, 2015:
http://www.theverge.com/2014/12/15/7393665/skype-translator-features
Wu, Xindong, Vipin Kumar, J. Ross Quinlan, Joydeep Ghosh, Qiang Yang, Hiroshi Motoda, Geoffrey J. McLachlan, Angus Ng, Bing Liu,
Philip S. Yu, Zhi-Hua Zhou, Michael Steinbach, David J. Hand, and Dan Steinberg, “Top 10 Algorithms in Data Mining,” Knowledge and
Information Systems, Volume 14, No. 1, December 2007, pp. 1–37.
30
Acknowledgments
The authors would like to acknowledge the participation and assistance of the panelists of the Web 3.0+ Criminal Justice Workshop. We
would also like to acknowledge the contributions of William Ford and Steve Schuetz of the National Institute of Justice. The authors also
acknowledge the valuable contributions of the two peer reviewers of the report, Edward Balkovich and John Davis, both of RAND.
The RAND Safety and Justice Program
The research reported here was conducted in the RAND Safety and Justice Program, which addresses all aspects of public safety and the criminal justice system, including violence, policing, corrections, courts and criminal law, substance abuse, occupational safety, and public integrity.
Program research is supported by government agencies, foundations, and the private sector. This program is part of RAND Justice, Infrastructure,
and Environment, a division of the RAND Corporation dedicated to improving policy and decisionmaking in a wide range of policy domains,
including civil and criminal justice, infrastructure protection and homeland security, transportation and energy policy, and environmental and
natural resource policy. Questions or comments about this report should be sent to the project leader, Brian A. Jackson (Brian_Jackson@rand.org).
For more information about the Safety and Justice Program, see www.rand.org/safety-justice or contact the director at sj@rand.org.
31
About the Authors
John S. Hollywood is a senior operations researcher at the RAND Corporation and a professor of policy analysis at Pardee RAND Graduate School. His principal focus is information systems research in support of improving security, ranging from crime prevention to terrorism
prevention to improving combat effectiveness. His recent research projects have included examinations of predictive policing, high priority
information technology needs for law enforcement, and U.S. effectiveness at foiling terror plots.
Dulani Woods is a data science practitioner adept at data acquisition, transformation, visualization, and analysis. He has a master’s degree in
agricultural economics (applied economics) from Purdue University. His master’s thesis was an economic analysis of organic and conventional
agriculture using the Rodale Institute’s Farming Systems Trial. He began his career as a Coast Guard Officer on afloat and ashore assignments
in Miami, Fla.; New London, Conn.; and Baltimore, Md.
Richard Silberglitt, senior physical scientist at the RAND Corporation, has worked in academia, government, and private industry for more
than 40 years, performing, evaluating, and managing research in materials, energy, nanotechnology, and other advanced technology areas.
He has lectured widely on emerging technologies in the United States and Asia, and he served as chairman of the International Advisory
Board of the APEC Center for Technology Foresight, in Bangkok, Thailand.
Brian A. Jackson is a senior physical scientist at the RAND Corporation and director of the RAND Safety and Justice Program. His research
focuses on criminal justice, homeland security, and terrorism preparedness. His areas of examination have included safety management in
large-scale emergency response operations, the equipment and technology needs of criminal justice agencies and emergency responders,
and the design of preparedness exercises.
About This Report
On behalf of the U.S. Department of Justice, National Institute of Justice (NIJ), the RAND Corporation, in partnership with
the Police Executive Research Forum (PERF), RTI International, and the University of Denver, is carrying out a research effort
to assess and prioritize technology and related needs across the criminal justice community. This initiative is a component of
the National Law Enforcement and Corrections Technology Center (NLECTC) System and is intended to support innovation
within the criminal justice enterprise.
This report is one product of that effort. It presents the results of the Web 3.0+ Criminal Justice Needs Workshop, a
group convened in September 2014 as part of the NIJ/NLECTC Priority Criminal Justice Needs Initiative to identify and
prioritize future criminal justice technology needs on emerging Internet technologies such as linked data, Semantic Web,
intelligent agents, and the Internet of Things. This report and the results it presents should be of interest to planners from
law enforcement departments, corrections agencies, and courts; research and operational criminal justice agencies at the
federal level; private-sector technology providers; and policymakers active in the criminal justice field.
This publication was made possible by Award Number 2013-MU-CX-K003, awarded by the National Institute
of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or
recommendations expressed in this publication are those of the author(s) and do not necessarily reflect those of
the Department of Justice.
Limited Print and Electronic Distribution Rights
This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property
is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited. Permission is given to
duplicate this document for personal use only, as long as it is unaltered and complete. Permission is required from RAND
to reproduce, or reuse in another form, any of our research documents for commercial use. For information on reprint
and linking permissions, please visit www.rand.org/pubs/permissions.html. For more information on this publication, visit
www.rand.org/t/rr928.
© Copyright 2015 RAND Corporation
www.rand.org
The RAND Corporation is a research organization that develops solutions to public policy challenges to help
make communities throughout the world safer and more secure, healthier and more prosperous. RAND is
nonprofit, nonpartisan, and committed to the public interest.
C O R P O R AT I O N
RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.
registered trademark.
R® is a
RR-928-NIJ