|
|
|
Department
of Energy
19901 Germantown Road
Germantown, MD 20874
October 3, 2000
WELCOME
Kurt Molholm, CENDI Chair, opened the meeting at 9:10 am. He thanked DOE for hosting the meeting.
Dr. Ed Oliver, Associate Director for Advanced Scientific Computing Research (ASCR), welcomed CENDI to DOE. He described the Advanced Scientific Research organization, which includes the Office of Scientific and Technical Information. There are four activities that touch every part of DOE: OSTI, scientific computing, SBIRs, and technology transfer. The activities track very closely to those of the National Science Foundation (NSF). OSTI is charged with capturing the products of all the research.
DOE is in the middle of the current budget discussions. While DOE received more money from the House/Senate Conference Committee’s mark-up than the President’s request (a 40 percent increase), OSTI essentially was budgeted the same amount as the previous year. Because OSTI is a separate line item, there is no flexibility to move money from the other ASCR activities. The budget increase for computer-related activities was a result of letters from university presidents and members of Congress to the House and Senate conference chairs indicating the importance of computing in the physical sciences.
Dr. Oliver acknowledged the recent workshop at the National Academy on building a future information infrastructure in the physical sciences as a significant event, which resulted in an excellent report. There are so many opportunities at OSTI but the budget has not provided for them. DOE is currently trying to get an information infrastructure together that will support these ideas. Dr. Oliver believes there is bipartisan interest in interagency collaboration, and CENDI is a perfect vehicle for doing this.
In terms of the information future, computing will not be the problem. Soon there will be 100 gigabit networks that can deal with the dissemination of new kinds of information, and the cost of such a network will be relatively cheap. However, we need to know how to manage the content.
Dr. Oliver believes that the connections between information, SBIR programs, and technology transfer have not been adequately exploited. There does not appear to have been any direct analysis of the impact of access to information and the success of technology transfer.
Discussion
Most agencies cannot directly identify the SBIR grantees or proposers that use their systems because the statistics are not kept at a granular enough level, due to privacy concerns.
Dr. Oliver was questioned about the relationships between his organization and the DOE CIO. He indicated that DOE is presently without a CIO. The previous relationships have varied from intimate to avoidance. The CIO actually relies on the ASCR for support in many technical activities related to scientific computing, but Dr. Oliver does not report to the CIO.
Dr. Walter Warnick, Director of OSTI, also welcomed the CENDI members to the Germantown complex. He gave a history of the department dating back to the discovery of fission in the 1930s. 1942 saw the first demonstration of a chain reaction under the direction of Enrico Fermi. The subsequent development of the atomic bomb was one of the most significant efforts in the history of science. With this effort came the creation of the DOE laboratories such as Oak Ridge, Hanford, and Los Alamos as well as the formation of the Atomic Energy Commission (AEC). Following the war, the Manhattan Project and the laboratory facilities were expanded beyond the development of the bomb to peaceful uses of nuclear technologies, such as in medicine, power generation, etc.
The Germantown facility was built in 1957. In 1974, the AEC was absorbed by the Energy Research and Development Administration (ERDA). The mission expanded within ERDA to go beyond nuclear energy to include fossil fuels and alternative fuels. DOE was established to include both ERDA and the Federal Energy Administration. The Nuclear Regulatory Commission (NRC) was separated out from AEC in 1974 because of the need to segregate the promotion of nuclear powers and ???? regulation. Now, many believe that the NRC regulation has been too tight. No new nuclear power plants have been built in the last 20 years. The current concerns focus on how to sustain existing nuclear plants, since many are approaching their original life expectancy.
Continuing its emphasis on Information Infrastructure, CENDI has been following two key activities, the assessment of public information dissemination, including the closure of NTIS being conducted by the National Commission on Library and Information Science (NCLIS), and the Information Technology Research (ITR) Program under the auspices of the National Science Foundation (NSF). The October program included updates on these two activities, as well as a demonstration of new services from DOE.
"Toward an Analysis of
Federal Information Policies: Comprehensive Assessment of Public Information
Dissemination Reforms"
Robert Willard, National
Commission on Library and Information Science (NCLIS)
NCLIS is celebrating its 30th anniversary this year. Mr. Willard believes that in its history, NCLIS can point to a few activities that are highlights and the current activity on the possible closure of the National Technical Information Service (NTIS) and assessment of public information dissemination policies is one of these highlights. This effort is directly in line with NCLIS’ mission to help Congress and the President to provide information to the American public.
Mr. Willard began by giving a brief chronology of the NCLIS work in this area. In August 1999, the Department of Commerce (DoC) proposed to close NTIS. In the Fall of 1999, the Senate and House Oversight committees held hearings regarding the closure, and in October 1999, NCLIS began an independent assessment of the impact of such a closure. This report was issued in March 2000 (www.nclis.gov/govt/ntis/ntis.html). It recommended that NTIS remain intact through FY2001 while further study is conducted to determine the impact of such a closure and that appropriate government funds be identified to cover the NTIS functions that are inherently governmental in nature rather than depend on cost recovery. The study also requested that Congress authorize NTIS to continue to offer other services on a cost-recovery basis in compliance with OMB (Office of Management and Budget) Circular A-130, provide funds to defray the one-time cost for NTIS to offer full service to the Federal Depository Library Program (FDLP), and restore NTIS to a satisfactory level of capacity, staff and service in the short-term.
The current situation is that legislation has not been introduced that would close NTIS. NTIS continues to operate with a surplus for FY2000. A hiring freeze has been imposed but NTIS recently got approval from DoC to supplement NTIS staff with contractors.
In June 2000, the Senate Commerce Committee requested additional study by NCLIS. The NTIS issue became part of a larger activity to look at the dissemination of public information. Senator John McCain, chair of the Senate Commerce Committee, asked for 1) a review of the government dissemination policies and identification of possible reforms, 2) a proposal of new or revised laws, rules, regulations, missions and policies in this regard, 3) suggestions for modernizing organization structures and functions, and 4) ideas for strengthening key components of the federal information dissemination infrastructure.
In July, the Senate Governmental Affairs Committee, chaired by Senator Joseph Lieberman, requested that a review of the Paperwork Reduction Act (PRA), which is up for reauthorization in 2001, be added to the study. This involves considering the viability of NTIS as a centralized, fully electronic repository for federal STI. In August 2000, Representatives Morella and Davis asked that the Government Accounting Office (GAO) study NTIS, in particular the size, age and demand for NTIS holdings, the extent of NTIS’ holdings from other sources, and federal compliance with the laws requiring deposit of STI.
The co-coordinators of the NCLIS study are consultants Woody Horton and Sarah Kadec.
NCLIS is extremely interested in interactive development so all information is quickly posted to its web site (www.nclis.gov/govt/assess/assess.html) in draft status. NCLIS is also interested in long-term access and preservation of all information related to this work, and plans to keep this information available.
The structure of the current activity relies heavily on the involvement of other interested parties, since there is a lot to accomplish by the December 15 deadline. There are four panels – reforms/updates to the NTIS Business Model, 2) federal agency clients and an evaluation of service bureau requirements, 3) a review of policies and practices for public access and dissemination from the standpoint of external users, and 4) a review of the private sector and public sector roles in government information. Over 100 people have agreed to work on this effort. The panels are working independently without commission members or staff in attendance.
The first panel, chaired by Peter Urbach formerly of NTIS, will analyze both short- and long-term alternatives. It will also propose new business models for NTIS that are consistent with NCLIS’ overall recommendations for changes in information dissemination policy.
The panel on federal agency needs (Panel 2) is chaired by Kurt Molholm (DTIC and CENDI). The goal is to identify the support needed by government agencies from an NTIS-like function in order to perform required information dissemination and internal missions. The panel is also looking for ways to increase interagency sharing of public domain information.
Miriam Drake of Georgia Tech is chairing the third panel on external user needs. The panel will identify and address diverse user needs in the Internet environment, especially for users with special needs and the disadvantaged.
The fourth panel is reviewing private/public sector roles. It is chaired by Wayne Kelley, former Superintendent of Documents and information executive in the commercial sector. The plan is to define and distinguish appropriate roles for access and dissemination of government information in the Internet environment. NCLIS plans to republish the 1982 NCLIS Report on "Public/Private Sector Interaction in Providing Information Services". Mr. Willard noted that this group is discussing the definition of government information and finding this a difficult challenge.
A Board of Experts in information technology, economics, librarianship, and legal matters has been formed to review the results of all the panel input. Members of the group have been selected for their expertise as well as for their lack of vested interest in the outcome. NCLIS plans to continue to coordinate with other agencies across the branches and with interagency groups such as FLICC and CENDI, and to continue the dialog with outside organizations such as library associations. NCLIS has asked members of the Board of Experts to provide white papers that would spur new ideas and perspectives throughout this process. Public comment continues to be solicited with regard to all these activities.
The schedule calls for completion of the panel reports, the white papers, and the information resources map by October 15, 2000. The Board of Experts will complete its review by November 15. The final report and recommendations will be sent to Congress and the President on December 15. After December 18, NCLIS will conduct briefings for various stakeholder groups. Communities will be able to comment on the report after the December 15, 2000, submission to Congress if they feel that they have not been well represented in the outcome.
Mr. Willard emphasized that the NCLIS Web site is a dynamic "Bulletin Board" that is being used to encourage public awareness. The Web site is being used as the main communications mechanisms, because the comment periods for posted documents are extremely short. He suggested that those interested in following the activities of the panels and NCLIS access the board often in order to keep up with the rapid developments.
Discussion
Mr. Willard was asked about other legislative initiatives that might result from this study. He is unsure what other legislation might come from this, other than the specific request to address the PRA reauthorization.
Mr. Molholm mentioned that his panel is developing a paper on life cycle management to identify the myths and challenge the paradigms, particularly relating to electronic environment. Mr. Molholm asked for comments on the report after it is made available from the NCLIS site. Another paper is planned on the categorization of information.
Mr. Willard was asked about the connection between this activity and the Lieberman-Thompson Web site on government information dissemination. He indicated that NCLIS sent a letter to Senators Thompson and Lieberman about the study.
The status of the Government Accounting Office (GAO) Study was discussed. Mr. Finch and Mr. Neuschatz indicated that the study is just beginning. GAO will be coming to NTIS in the near future to begin the analysis, which includes NTIS’ audiences, the depth and breadth of the collection, and other places users might go to obtain this information.
"The NSF Information
Technology Research Program: Agency Requirements and Projects"
Dr. George Strawn, National
Science Foundation
Fundamental long-term research programs in information technology are extremely important, especially at times when new opportunities are exploding. Dr. Strawn cited Douglas Robertson’s book, The New Renaissance, which postulates that renaissances in human history are associated with information explosions such as spoken language, written language, printing, and, most recently, electronic computing. Similarly, Michael Hobart and Zachary Schiffman, in Information Ages, have identified three ages of information: the age of literacy, the age of numeracy or mathematically driven science, and the age of computeracy. The latter is reflected in the development and use of computers to simulate and model complex systems. In order to achieve the latter age, it is necessary to have longer-term development of information technologies at the fundamental, rather than the application, levels.
The Information Technology Research Program (http://www.itr.nsf.gov) is an outgrowth of the 1999 PITAC (President’s Information Technology Advisory Committee) report. This group formed in 1997 and was made up of 26 representatives from academia, industry and research organizations. The PITAC concluded that while Information Technology (IT) will transform the way we live in the 21st Century, the Federal information technology R&D investment is inadequate. In constant dollars, the support has been flat or declining for the past ten years. The PITAC also found that federal IT R&D is too heavily focused on near-term, mission-oriented problems and not long term IT research. The PITAC recommended that the Federal government create a strategic initiative in long-term IT. It recommended that funding be increased by $1.37B per year by FY2004, that visionary and high-risk research be encouraged, and that diverse modes of research support be used and the duration of projects increased.
Building a Federal IT program of value in the 21st century will require new management strategies, new modes of research support, and new implementation strategies. The PITAC also recommended that NSF take a leadership role in basic IT. Extension of the HPCC program to coordination of all projects was also requested. While the major focus is to be on long-term research, funding collaboration with applications to drive IT was also identified as a strategic goal. The PITAC also recommended the funding of centers for "Expeditions into the 21st Century."
In its report, the PITAC outlined 10 ways that IT will transform our society -- the way we: communicate, deal with information, learn, practice health care, design and build things, conduct research, understand the environment, conduct government, and the nature of commerce and work.
Four areas were identified as needing major emphasis during this period – software, scaleable infrastructures, high-end computing, and the socioeconomic impacts of IT.
In the area of software, we know how to develop good software, but we don’t know how to make it both good and affordable. The current methods include fast development in the commercial environment that results in situations where users are often providing the beta test results. Development of hardware moves very quickly but the methods for quality software development have lagged behind.
In the area of network infrastructure, we need a scaleable information infrastructure. The concept will no longer be limited to just people on the Internet, but computers and the Internet involved in all kinds of appliances and tools. However, we aren’t able to model the Internet in its current stage, let alone one that is even more pervasive.
High-end computing is necessary to attack the issues of complex systems for ecological analysis, biochemical analysis, etc. Scientists are anticipating sustained petaFLOP/s systems. As advances are made in software, high-end computing and the infrastructure, the socioeconomic impact is increasingly important. The digital divide exists and is growing within key groups such as older Americans and ethnic groups.
In response to the PITAC report, President Clinton and Vice President Gore proposed the Information Technology for the 21st Century (IT2) initiative with funding of $366M FY00 or an increase for NSF of $146M. Congressman Sensenbrenner introduced the Networking and Information Technology Research and Development Act that passed the House, but not the Senate. It called for a five-year authorization of $326M in FY00, or an increase for NSF of $241M. Congress actually passed a federal budget that included a $126M increase for NSF in the area of "ITR" (which replaced the IT2 name).
NSF has conducted several major activities related to ITR in FY2000. This includes a $36M award to the Pittsburgh Supercomputing Center for the acquisition of a terascale (5 Tera/flop) computing system for science, engineering, and the nation. It also includes the $90M program to address the PITAC research recommendations. This is a crosscutting program to include all areas of science and engineering. Large grants of $3M a year or more for up to five years and small grants of $500,000 or more for three years were included.
The eight areas of research align closely with the four major areas defined by the PITAC. They include advanced computational science, human-computer interface design, information management, IT education and the workforce, revolutionary computing, scalable information infrastructure, social and economic implications of IT, and software.
The advanced computational science area includes algorithms, software, and systems applicable to science and engineering. These include numerical methods, software libraries, data handling, scientific visualization, etc. This area strongly encourages multidisciplinary research.
The human-computer interface area seeks to address the understanding of human perceptual, cognitive, and social abilities and their relation to interface design. The goal is to improve our ability to use computer in areas such as language processing, computer perception, and visualization. The use of computer should be extended through "assistive interfaces" and collaborative uses of systems.
The Information Management research will seek to increase the utility and scope of online information, content and access. Examples include the management of new kinds, media and genres of data, long-term digital archiving, data fusion across different sources, and the linking of NSF’s ITR activity with research being conducted in other sciences and engineering.
Education and the Workforce seeks to identify new methods for educating people in IT at all age and educational levels, and also to increase IT literacy and skills among the general student and public populations.
Revolutionary computing involves the development of potential new technologies that do not share the physical limitations of CMOS and that identify different ways to computer, perhaps in even smaller environments. Examples include DNA computing, optical computer and quantum computing.
The Scalable Information Infrastructure addresses fundamental aspects of networking including the need to simulate and emulate very large networks, further development of network technologies such as optical and wireless, and network security and privacy tools.
The Social and Economic Implications of IT seeks to understand the underlying processes by which IT influences and transforms society in areas such as the economy, commerce, and distributed work environments.
The Software research area seeks to improve the technology for software creation with the goal of making software more reliable, robust and adaptable. Other areas include software security and metrics.
To date, 140 awards have been made of $150K per year for three years. About 60 awards averaging $1M per year and up for five years were announced.
Looking to FY2001 and beyond, the PITAC has been re-authorized until February 2001. The PITAC continues to review and report on IT matters including the cross-agency Next Generation Internet program, the digital divide, and the ITR program at NSF. There are a variety of new subcommittees preparing recommendations in areas such as digital libraries, digital government, open software and tele-health.
In terms of federal budgets, Dr. Strawn acknowledged a strong bipartisan support for basic IT research. However, he noted that IT research has to be paid for in the context of the entire federal budget. NSF has proposed a budget of $45M for the Terascale Computer System and $100M increase for CISE for ITR. An additional program announcement was recently released.
The areas of research for FY2001 remain very similar to those for FY2000. They include system design and implementation, people and social group interactions with computers and the infrastructure, information management, applications of computers to science and engineering, and scalable information infrastructures for pervasive computing and access. These areas encompass the previous areas identified by PITAC and the FY2000 ITR program areas. There have been additional application emphases in the areas of environmental informatics, geoscience modeling, biological informatics and information in math and the physical sciences. The modes of funding for FY2001 will include three sizes of grants, single-PI projects, small groups, and large groups. This may change depending on the quality of the proposals.
Dr. Strawn ended with a request for reviewers from among the federal science domain community. The reviewers are key to NSF’s peer-reviewed programs.
Discussion
Dr. Strawn was asked about the connection between these NSF programs and similar activities in other countries. He indicated that the high performance and infrastructure areas are part of the Global High Performance Network effort. However, he believes that the US is still about five years ahead of other developed countries, particularly in computer penetration in schools and homes. He attributes much of this to the liberalization of the telecommunications industry.
Dr. Strawn noted that intellectual property issues, particularly with universities, have made the line fuzzier between short-term applied research and longer range basic research. However, the industries, especially those represented on the PITAC, feel that the government needs to be involved in long term fundamental research, because the private sector cannot support this.
"DOE Scientific and Technical
Information (STI) in a Period of Profound and Rapid Change"
Dr. Walter Warnick and
Vincent Dattoria, DOE/Office of Scientific and Technical Information
This briefing was originally presented to the PITAC Digital Library Panel on September 19, 2000. It was triggered by the NAS Workshop Report on a Future Information Infrastructure for the Physical Sciences, which is available at >www.osti.gov/physicalsciences. This is a vision statement. What is needed next is implementation strategy for an information infrastructure for the physical sciences.
The goal is a common knowledge base for comprehensive access for use and reuse of physical science information. This requires that the approach be interagency, because the physical science research is conducted across a number of agencies. In addition, it must work off the strengths of the individual agency missions. For example, NSF has carved out the area of education with the National Science, Mathematics, Engineering, and Technology Education Digital Library (NSDL), while DOE has emphasized content for researchers in particular areas. Second, the approach must include multiple document types. Dr. Warnick introduced a trilogy of scientific and technical information including the gray literature, journals and preprints. Each document type collection is managed separately, because there are different means of acquisitions, different formats, and different degrees of reliability.
Dr. Warnick showed some early successes at cross agency distributed searching. Using a distributed search engine called Distributed Explorer, OSTI has developed services to search web-based gray literature and R&D project descriptions across agencies. The GrayLit Network includes 100,000+ technical reports that are publicly available from DTIC, NASA, EPA and DOE. OSTI plans to connect to NTIS for distribution of the paper documents upon request, now that DOE has implemented a Persistent URL (PURL) system.
The Federal R&D Projects service includes approximately 220,000 R&D summaries from DOE, NIH and NSF. Again, a cross-database search engine was developed to search across the three R&D databases and to provide a single set of results.
Journal articles are made available through PUBSCIENCE, which aggregates journal articles from multiple publishers under one search engine. Links are provided to the full text of material on the publisher’s site.
The components of OSTI’s digital library for energy science can serve as the building blocks for part of the physical sciences infrastructure. Additional building blocks would be provided by the other science agencies. The glue that holds the blocks together is the distributed search technology.
Discussion
Mr. Molholm asked Dr. Warnick how the Federal R&D Projects service differs from RADIUS. Dr. Warnick mentioned that the product is online, free and available. He also indicated that there is a basic problem with all project summary systems, because it is generally impossible to balance the actual budgets against those in the project system. Despite this problem, the DOE R&D Council agreed that OSTI should go ahead with the R&D Projects service. DOE has embarked on a new $37M project to develop a cradle-to-grave tracking system for R&D, which will help the data quality.
Mr. Molholm asked if there have been any ramifications from having so much information available to the international community. Dr. Warnick mentioned that international exchanges are important for DOE and include the 100 countries within the International Nuclear Information System (INIS) and the 20 countries in the Energy Technology and Data Exchange program. Mr. Molholm asked if the Internet has changed the balance between U.S. and foreign information. Mr. Smith mentioned that open communication and international exchange is becoming more acceptable in Congress. Mr. Molholm mentioned that DoD is looking at membership in TechWatch. This is a cooperative effort including the UK, Canada, New Zealand and Australia to identify and investigate early technology developments.
The similarities and differences between this system and FirstGov (www.firstgov.gov) were discussed. The primary difference is that FirstGov does not search within the databases themselves, while distributed search technologies do. The group would like to know more about the plans for FirstGov.
There were several questions about how the Distributed Explorer operates. Basically, the software creates a series of parallel search strategies behind the scenes. However, the strategies are currently at the lowest common denominator and do not take advantage of the nuances of fielded searching and special search parameters available from some native search engines. This, of course, impacts the precision of the search. Dr. Warnick indicated that this method could be viewed as a "first cut" to identify the most important systems to search, since users are pointed to the originating system where they can do additional searching in the more powerful, native search language.