Dec 31 2009

Burying Bureaucracy

Departmentalized services go the way of the horse-drawn buggy as government agencies team up to share data.

When the nation's
capital was plagued
by a series of deadly
shootings in the
fall of 2002, the
sniper made a phone
call to police and let
slip a sliver of information. Investigators
followed that lead to Alabama, and from
there proceeded to tie together a series of
crimes and identify a suspect.

While the sniper case "lead" was actually
buried in a criminal database, it wasn't
found through police investigation of
databases. The reason? Until recently, aside
from sharing rap sheets and other basic
criminal information housed in national
databases, criminal justice agencies
nationwide gathered leads and built cases
using mostly local data and resources.

That fact has led many to wonder
whether a national system designed to
let agencies share detailed information
could have led police to the sniper sooner.
While that question is unanswerable, the
Department of Justice and the Department
of Homeland Security are leading a charge
to see that this situation never happens
again. To that end, criminal justice agencies
across the country are building processes to
link their databases and share information.

Law enforcement isn't alone in its
quest. From the National Institutes of
Health to the State Department to the
General Services Administration (GSA),
information-sharing projects are helping
federal agencies reduce bureaucracy and
meet their mandates more effectively
and efficiently.

"It's breaking down the walls of
government," says Amy Santenello, an
analyst at META Group, an IT research
firm in Stamford, Conn.

Have Data, Will Travel

Sharing data and resources is at the heart
of the president's Quicksilver E-Gov
initiatives, which are aimed at making
government services more timely, useful
and cost-effective. As an example, in April
2002, GSA launched its e Travel Quicksilver
initiative, which centralizes travel
management services for the federal
government. (The only exception is the
Defense Department, which operates its
own travel system.)

All federal agencies had to start work on
implementing the Web-based travel system
by January 2004, and they must complete
deployment by Sept. 30, 2006, says Tim
Burke, GSA project manager for e Travel. In
the first six months of 2004, more than 200
Federal Aviation Administration travelers
booked, authorized and created vouchers
for foreign and domestic trips on eTravel.

"Data sharing is absolutely critical in
today's environment of heightened security
and e-government services," says Karen
Evans, administrator of the Office of
Electronic Government and Information
Technology at the Office of Management
and Budget (OMB). "The notification of
terrorist activities, the management of the
nation's military, online citizen access to
government programs and the use of
e-government programs all require the
sharing of data in an unprecedented way."

OMB's Federal Enterprise Architecture
(FEA) program provides a common
language through a set of reference models
that establish a structure for classifying
and organizing complex information
throughout the government, Evans adds.
The goal is to support the sharing of
information between federal, state, local
and tribal organizations.

Another example of federal data-sharing
initiatives is the Department of Interior's
Recreation One-Stop program, which is
intended to improve access to recreation-related information that's generated by the
various levels of government. Other
initiatives include the Department of Health
and Human Services' Federal Health
Architecture program. It consists of a
partnership among 12 federal agencies—including the Veterans Administration,
the Department of Defense and the
Environmental Protection Agency—to
leverage an effective, efficient mechanism
for making sound decisions that impact the
nation's health.

One key to any successful data-sharing
project is to develop common processes
among agencies with different procedures,
technological capabilities, laws and
budgets. Privacy and security laws that
govern federal agencies can conflict with
laws in some states, making the quest for
common ground a difficult goal.

"The problem isn't technology," explains
Tom Kooy, president of Justice Information
Sharing Professionals, an association
of criminal justice practitioners that
spearheads data-sharing initiatives around
the nation. "The problem is all the different
lines of business and different boundaries
around management and budget."

The owners of the data and the business
processes must resolve these thorny issues
before technology can do its job and let data
flow freely between organizations.

Bridging Jurisdictions

The concept of sharing information isn't
new. About four years ago, the Justice
Department, under the aegis of the Global
Justice Information Sharing initiative,
brought together representatives from
local, state and tribal law enforcement
groups to find ways to build bridges
between agencies. Since then, this global
initiative has been convening working
groups on intelligence, data practices,
technology infrastructure and Web services
to agree on national standards, such as a
common criminal-justice dictionary and a
justice-specific Extensible Markup Language
data model, says Kooy, who served as a
group representative from Minnesota,
where he helped develop CriMNet, the
state's criminal justice network.

In August 2003, Justice took Global's
work a step further by initiating the Law
Enforcement Information Sharing (LEIS)
initiative, a strategy to link the nation's
18,000 criminal justice agencies—everything from rural police departments to
Justice Department agencies, including the
FBI and the Drug Enforcement Agency.
The goal was to share knowledge about
crime patterns and details, such as crimes
committed with a particular weapon,
vehicle or modus operandi.

Still in the planning stages, LEIS is
designed for phased implementation. Some
elements could be ready to deploy next year,
but it will likely be a constant evolution.

Privacy is a major issue for LEIS. Federal
law enforcement agencies are subject to the
U.S. Privacy Act, which lets them share
information they collect on people, but it
must be as accurate as possible and those
they share it with must show just cause for
viewing the information. Federal agencies
must also record and report on those
accessing its information.

States, though not compelled by the
Privacy Act, are governed by their own
privacy laws that often compete or conflict
with federal laws. For that reason, LEIS
collects only criminal information that
exists within the justice data community
rather than seeking out new data sources.

In terms of security, 80 percent of the
design will be dictated by laws and
regulations. The rest will derive from
agreements between participating agencies.

Identifying Open Standards

Although there are existing information-sharing systems, such as the Regional
Information Sharing System and the
National Incident-Based Reporting System,
many are antiquated and contain data fields
that are inconsistent with information
gathered from state and local agencies. So
adding data to these databases can be a
frustrating experience.

"It's a lot of rewriting or recoding of
source crime data, and you're trying to
pigeonhole crime types into a set of broad
categories," says Kooy. Thus, compliance
with such systems has been minimal.

These hurdles impede efforts like LEIS,
so the project has sought to identify
common, open standards, so agencies
can choose whether or not to participate.

The system will be built as a distributed
architecture that will tap into and index
local databases, so investigators will be able
to reach the original source of the data to get
complete, up-to-date records. Ideally, it will
be so valuable that agencies will volunteer
to participate. "You want to build a field of
dreams—something useful and reasonable
to local agencies—and let them build on it
with their own investment," Kooy says.


The goal of the National Cancer Institute (NCI) is a
lofty one: End the suffering and deaths from cancer
by 2015.

In labs nationwide, research teams spend
countless hours and dollars working toward that
goal, but have had no easy way to pool their tools,
research and brainpower. Technology will soon help
NCI clear that hurdle.

"We're bringing information technology to the
war on cancer," says Ken Buetow, director of NCI's
Center for Bioinformatics.

The new weapon is NCI's Cancer Biomedical
Informatics Grid (caBIG), a common unifying IT
platform that joins cancer research communities
to let them share infrastructure, applications,
processes, information and the products of their

Announced in July 2003, work on the grid kicked
off in February with pilots at 50 NCI-designated
cancer centers. The grid features a combination of
information and infrastructure standards.

There is no central data warehouse. Instead,
there's a federated model with local institutions
connecting over the Internet. The standards cover
issues such as security and data structure. It's a free
open source system, so anyone who complies with
the standards can join, including researchers,
clinicians and patients.

"The scientists can focus on innovation, not just
rebuilding and redeploying basic infrastructure and
basic tools," Buetow says. "You get the power of
collective insight—the utility of not just a single or
small number of bright imaginations looking at
a particular data set, but the entire research

CaBIG has already pulled together an inventory
of biomedical tools and data sets that weren't
previously accessible to the cancer community at
large. Researchers share tools to examine the
biological circuitry in humans to help treat cancer as
a systemic disease. They can catalog and test for
genes and proteins, study how they interact within
the human circuitry, then analyze how therapeutic
approaches alter them, Buetow explains.

"They can consider all sorts of explorations of
data that they couldn't contemplate before because
the effort of doing just one of them would be
all-consuming," he says. "I think it really has the
capacity to transform cancer research, and put us
well on our path to eradicating suffering and death
due to cancer."


Agreeing to share data is
the first step in building a
multi-agency data warehouse.
But before the building
process actually starts,
several questions must be

1. What information will be
collected? It should be
extensive enough to be
useful, but not so far-reaching
that it violates privacy laws.

2. Will members have access
to each other's databases,
or will a Web service provide
a centralized index of

3. What attributes should the
system include? Consider
all aspects of the system,
such as chat capabilities,
system redundancy and
two-level authentication.

4. Should participation in
the system be mandatory
or should agencies have
the option of signing a
memorandum of

5. Which agency will be
responsible for ongoing
development and

6. Does government
already have any of the
system building blocks in
place? Other agencies may
already have done the work
to create a secure intranet,
for example, and that can
save you from reinventing
the wheel.

7. Who can have access to
the system without
violating privacy laws?