CENDI PRINCIPALS AND ALTERNATES MEETING

National Library of Medicine

Bethesda , MD

April 6, 2004



MINUTES

E-Government and Integration Architectures
What's Ahead for E-Government and Federal Information Policy
The Information Infrastructure in DHS: Strategies and Plans
NLM Showcase
SeniorHealth.gov and the Talking Web
ACSI User Satisfaction Surveys: MEDLINEPlus and Science.gov Implications

Welcome

Kent Smith, Chair of CENDI, opened the meeting at 9:00 am. He introduced Dr. Donald Lindberg, Director of the National Library of Medicine. Dr. Lindberg welcomed the CENDI members. He expressed his pleasure that CENDI has continued to be a vital force for agency interaction. It has grown and continues to be a valuable forum. He invited participants to visit the special exhibit in the NLM Atrium called “Changing the Face of Medicine: Women in Medicine,” which highlights the accomplishments of women who now comprise over 50 percent of the students in U.S. medical schools. Kent Smith continued the introductions by formally congratulating Karen Evans and Bonnie Carroll on receiving Federal 100 Awards. In addition to the regular award, Ms. Evans received the prestigious Eagle Award.

E-GOVERNMENT AND INTEGRATION ARCHITECTURES

“What’s Ahead for E-Government and Federal Information Policy”
Karen Evans, OMB Administrator for E-government and IT, Office of Management and Budget

The President’s Management Agenda focused on getting control of the $60 billion information technology (IT) budget to ensure that the results are citizen-centric. Ms. Evans discussed the achievements that have been made and the process to determine where joint procurement or procurement of consistent IT options could result in better prices from vendors. The FY04-05 IT budget has received a significant increase with a one percent overall growth, considering that the budget deficit is to be cut in half in the next five years. The homeland security related budget is for all agencies, excluding the Department of Commerce. The language is still being worked out with Congress so the numbers are not yet available.

OMB’s major focus is to use tools, such as overall portfolio management, in relation to the budget process to break down the silos across agencies. What they are doing in the Form 300 is related to the PART. The goal is to determine if IT will help an agency achieve its goals and then identify how to measure this value. This in turn will add overall value to the programs. In the FY04-05 budget, there is more than $10 billion in new IT investments in lines of business such as financial management, human resources management, grants management, and federal health architecture. This new money will be directed by OMB in budget guidance to be redirected into the central systems that are cross-cutting agencies. There is a lot of money to be saved by working together. OMB is trying to get the analysis lined up in the FY06 cycle.

Another tool is the budget process itself. New development dollars were held over until it was determined if the dollars could be targeted for a cross agency initiative. They are not looking to do this as a tax on the agencies; rather, the steady state dollars are intended to allow the agencies to keep operating while new development is managed differently.

Ms. Evans also discussed the status of the 25 Quicksilver projects. Some of them are mature and will be moving into an “operational” environment. Utilization plans are being developed to determine what it means to be 100 percent completed. The projects may be complete technologically but incomplete in terms of utilization by their target audience. For example, Grants.gov needs to move to a fee-for-service model with dollars coming from service agreements with the agencies. All granting opportunities need to be listed on grants.gov and agency-specific application should be eliminated. In the original Quicksilver initiatives, the metrics are output oriented rather than outcome oriented. It is hoped this will change over time. IRSfreefile has developed good metrics as part of its utilization plan.

Several new Quicksilver-like groups began at the end of FY03. Grants management and human resources management were added. There are two initiatives within the Department of Homeland Security. The first is Safecom Wireless, which is standards and architecture driven. The second is disaster management, which is seeking to identify the basic level of service that first responders need and the vendors who are in this space.

The Forum of Chief Information Security Officers has replaced the Critical Infrastructure group. Instructions for critical infrastructure plans (HSPD7) are coming out soon; the plans are expected to OMB by July 1, 2004.

The Data Reference Model (DRM), under the guidance of Jonathan Womer and Bob Haycock, is in the official review stage and will be released for agency comment soon. The DRM will not be used for this year’s budget cycle. Updates to information dissemination policies will come through A-11, which is out for comment to the CIOs, rather than A-130. Ms. Evans expects minimal changes to A-130.

USA Services is part of the revised A-11. It is set up with multiple channels and Firstgov is one of them along with phone and e-mail. The Federal Citizen Information Center (FCIC) in Pueblo, Colorado, has been absorbed under USA Services. USA Services supports a basic level of service that is free to all agencies. They are investigating 800 numbers and call centers. Tier 2 and 3 responses will not be handled by USA Services but rather by the agencies. The Department of Commerce has sent out a “dear colleague” letter telling other agency heads that they are using USA Services.

There are a variety of barriers that need to be eliminated in order to have agencies achieve “green” status and move E-government forward. The result will be breakthrough performance that should free up money for other initiatives. Breakthrough Performance is defined as innovation in management practices and leadership in technology adaptations (allowing agencies to innovate instead of doing business as usual).

“The Information Infrastructure in DHS: Strategies and Plans”
Lee Holcomb, Chief Technology Officer, Office of the CIO, Department of Homeland Security

Three months ago, the Department of Homeland Security completed its vision and mission statements. A key phrase in the statement is “preserving our freedoms.” They are trying to balance freedom with security.

The DHS mission can be broken down into the following “value chains”: understanding the threats, prevention, protection of critical infrastructure, response, recovery in the aftermath, and service to the public with organizational excellence. A number of goals were laid out in the annual address by Secretary Ridge. The number one goal is information sharing. However, this is a daunting task involving both horizontal and vertical sharing. Mr. Holcomb believes the goal aligns directly with the interests and capabilities of CENDI.

Key to success in this area is to build communities of interest and tools for collaboration. Their approaches are similar to DoD in “pushing back the borders” by dealing with threats before they come near the boundaries. “Power to the edge” in their environment means putting computational power and intelligence at the borders.

In the midst of doing this, it is important to balance security and privacy. To help in this regard, the Department has brought in privacy officers from the private sector with a variety of different views to help them form their policies and activities.

In the sphere of information management, it is important to capture information once. A database of trusted information must be built that can help individuals differentiate threat from non threat. Hewlett Packard has referred to DHS as the “mother of all integration” efforts. They are trying to rationalize systems that are in place and then move toward one network. However, not everything can be done at once. A single database of names is now used by visa and border screeners. A single nomination process now takes place when determining a name that will be entered into the database, eventually a single database will exist. Authentication and governance structures must be in place too.

The Homeland Secure Data Network (HSDN) is a secret collateral network like SIPRNET. DoD is concerned about extending their SIPRNET so it will be developed as a peer network with DISA and others including the Department of Justice, (DOJ), Department of Energy (DOE), and the Department of State. Smaller agencies like the Department of Agriculture (USDA) could run on the HSDN.

In order to support information sharing at various government levels, they are pushing out connectivity. Secure video conferencing is now available with 56 states and territories. It is possible to communicate with chat or instant messaging. There are shared spaces with local officials that are secure.

A major challenge when dealing with state and local governments is the diversity of interfaces. Web-based services and other technologies will help. For example, DHS Info is being rolled out to 6000 participants in Dallas/Fort Worth. It is a local portal to create communities. Auto speed dialers are used to support alerts and warning. This model will be replicated elsewhere. DHS is working on this issue with the relevant state organizations such as the National Association of State Energy Officials (NASEO) and the National Association of Counties (NACO).

Identity management is a critical technology. The US Visit System will integrate more than 30 systems into a common dissemination system. Smart cards will be deployed to approximately 45,000 trusted employees next year.

Metadata is an issue. A metadata working group has been established within DHS. They are working with the Intelligence Community’s Metadata Group. A day-long workshop is being held between the IC Metadata Working Group and the XML Working Group to determine how to integrate across the Web and across communities. The Department is also involved in the Global XML Group and XML Standards for First Responders. They want to ensure compliance with basic standards like the Dublin Core. The Department has joined OASIS where they can work in the commercial arena. Metadata may be an area of joint interest between DHS and CENDI.

Action Item: Charles Bradsher and Gail Hodge will follow up with John Dickert at DTIC regard the work of the Intelligence Community Metadata Group.

Mr. Holcomb reported that the Department’s Online Library function is just beginning. They established a single DHS portal in 23 days using the Army Knowledge Online as the model. All agencies involved in the merger to DHS now have a common intranet. There is also an extranet version.

The Department has two new components: the Information Analysis and Infrastructure Protection (IAIP) and Science and Technology Directorate. The Homeland Security Advanced Research Projects Agency (HSARPA) is the DHS version of DARPA. Their strategy isn’t worked out yet, and strategic partnerships are being sought. Neither one of these components is under the Chief Technology Office but Mr. Holcomb indicated his willingness to bring interested parties from CENDI together with the appropriate individuals in DHS.

Action Item: Secretariat will contact Mr. Holcomb to arrange a meeting with representatives from DHS, including key individuals from the Online Librarian function, the Science and Technology Directorate, the IAIP, and CENDI. NTIS, DTIC, DOE, NAL, and GPO expressed interest in attending for CENDI.

NLM SHOWCASE

“SeniorHealth.gov and the Talking Web”
Dr. Simon Liu and Joyce Backus, National Library of Medicine

NLM, the National Institute for Aging, and others have developed SeniorHealth.gov, a web site that focuses on senior health issues. It was originally launched in October 2003.

Extensive usability testing was conducted. Focus groups were held at a number of senior citizens’ centers, with seniors of varying capabilities. A national group was sent homework and then a teleconference was conducted. These ranged from simple changes such as large buttons and fonts to more complicated issues surrounding hearing impairments.

Over 17 percent of the U.S. population over 45, or approximately 16 million people, have some kind of vision impairment, with approximately 180 million worldwide, according to the World Health Organization. These statistics will grow as the population ages and seniors live longer.

A number of technologies such as screen magnifiers, screen readers, speech recognition, and Braille devices can be used to improve accessibility for the visually impaired. However, these technologies require software on the users’ machine and some of them can cost as much as $4000 for an individual user.

NLM’s approach is to use large text, high contrast, and text narration. The two features that were demonstrated are the ability to increase the character size of the content. The second feature is the ability to have the web site converted to speech using computerized text to speech conversion. The narration files reside on the NLM server and the only requirement for the user is a sound card. This implementation also takes into consideration that many of the seniors are not technically savvy and they may not have the necessary computer equipment to run special plug-ins or clients. The learning curve is low. The user can support the degree of magnification and request these special features only if needed. Also, it works on a variety of browsers and platforms (NLM identified over 40 different combinations of browsers and platforms.) Performance was a potential problem with the server-based application. However, with AT&T Bell Labs Text Processing, they have successfully balanced the need for compression with quality of the audio files.

Domain experts use a content management system to create objects that will become part of the web site. Once the text is created, it is submitted to the database, converted to speech, and then the audio file is compressed. The compressed audio file is stored with the text. This allows for easy repurposing and consistency in the end product. There is a need in the near future to fine tune the system by enhancing the dictionary and the grammar, particularly with regard to disease and drug names.

“ACSI User Satisfaction Surveys: MEDLINEPlus and Science.gov Implications”
Dr. Fred Wood, National Library of Medicine

NLM has been using increasingly complex usability testing mechanisms. The information gained from such testing provides user feedback. It is a way for the agency to learn more about the users, including an understanding of the impacts of the information. You can learn how to improve the Web site, and how to gauge the impact of web site changes. Monitoring trends over time is becoming easier to do.

Three years ago, NLM took the lead in the development of user metrics with the workshop that was sponsored by CENDI. Part of the rationale for that workshop was to introduce the idea of integrated multiple approaches.

Online user surveys are in transition. They started out in phase one with just simple questions posted to the web. In the second phase, snapshot and randomized surveys were conducted. Approximately 2000-4000 surveys were received over two weeks. Limited benchmarking was available. Dr. Wood believes that they are embarking on the third generation of user survey which involves continuous sampling. There is a smaller responder size each month but the total in a year is about the same as in phase two. By using the American Customer Satisfaction Index (ACSI), OMB clearance can be expedited and the surveys can be approved in a matter of days. ACSI was developed by the University of Michigan business school.

Dr. Wood reported on the outcome of NLM’s recent evaluation of the ASCI Survey. (NLM has co-authored a paper to be published in the IEE’s IT Pro magazine in the near future.) NLM is piloting five web sites using ACSI. Two sites are from other NIH institutes and three are from NLM, including MEDLINEPlus.

With the ACSI, it is possible to add custom questions to the survey and to modify the wording of existing questions. NLM added four to six custom questions.

NLM believes that this approach compares favorably to the snapshot surveys and gives more information, more often. The cost is $2000-$2500 per year, compared to $2500-$2800 for only one snapshot survey per year. There is also the opportunity for benchmarking with others. Based on the pilot test, NLM compares favorably with a cross section of portal sites in the private sector. The evaluation team will likely recommend ACSI for future use by both NLM and NIH.

Five to ten percent of the people given the opportunity will respond to these surveys. This seems to be a consistent response rate, and there doesn’t seem to be any way to improve this rate without using cookies. This approach is probably biased towards savvy users and frequent users. “Pop-up blockers” may become a problem in the future. The Claes Fornell International (CFI) Group, developer of the ACSI, is working on a technology solution. There is also the issue of the “accidental user,” which the National Cancer Institute is tackling. Many “accidental users” come from search engines. Dr. Wood believes that the results are representative for knowledgeable users but they may not be equally valid for infrequent users.

There are approximately 44 agencies using the ACSI through the Federal Consulting Group at the Department of Treasury. Dr. Wood noted that NAL, NASA, and NARA already have agreements with the Federal Consulting Group. This may have implications for future Science.gov usability testing.

Briefings by Observers

Brand Neimann from EPA described the new Component Technology interagency group. It has been organized to foster component-based architectures. He distributed a brochure. The meetings are open and all are invited. There is a Best Practices Committee which is dealing with knowledge management. They have been asked to start a Semantic Interoperability Group as a new Community of Practice. There are new interoperability open standards and seven have been awarded.

Return to Minutes Archive