|
CENDI PRINCIPALS AND ALTERNATES MEETING |
U.S. Geological Survey/Biological Resources Division
Reston, VA
October 1, 1998
WELCOME
Tom Pedtke, CENDI Chair, began the meeting at 9:10 a.m. He thanked USGS/BRD for hosting the meeting. Introductions were made. The agenda was modified, but the minutes reflect the original order of agenda items.
BUILDING NEW PROGRAMS AND NEW TOOLS
National Commission on Libraries and Information Science: Building a New
Organization and Planning for the Future
Bob Willard, Executive Director, NCLIS
The National Commission on Libraries and Information Science, (NCLIS) was established in 1970 as a result of recommendations from President Johnson's National Advisory Commission on Libraries. The organic legislation recognized the value of information in meeting national goals. It called for senior level citizens from around the country to advise both the Executive and Legislative Branches on issues relating to libraries and information. An action plan, completed in January, 1998, has many of the same themes as the early action plan of the 1970's. While technologies have changed, many of the same issues exist for the use of information and public access to it.
NCLIS has three major areas of focus: 1) national and international infrastructure, 2) federal aid to libraries, and 3) access to government information.
National and International Infrastructure
In this area, NCLIS pays for and conducts surveys on the penetration of the Internet into public libraries. In 1994, the penetration was about 21 percent but the Internet was only available to library staff. In 1995, the study focused on cost. In 1996, the 1994 survey was repeated and the penetration was 42 percent. By 1997, it was 71 percent, and in 1998, the penetration is 83 percent. The full text of the latest report is available from the NCLIS web site (http://www.nclis.gov) along with a brief Powerpoint presentation of the highlights.
In addition to looking at general statements of use, the studies considered Aeffective" public access, including number of terminals, speed of access, and the graphic capabilities of the workstations used. If these effectiveness measures are taken into consideration, the number of libraries with adequate access (a graphics terminal with a minimum 56K connection) is less than half of the number of libraries with Internet access. Two-thirds of the libraries had two or less terminals with Internet access.
Another Internet issue is that of keeping undesirable sites from young children. While this has been a big political issue, NCLIS studies show that it is already being addressed by most libraries. Approximately 98 percent have software filters or policies in place. A similar issue involves child predators who make contact via the Internet. A report Internet and Kids: The Promise and the Peril will be prepared following a hearing to be held on November 10, 1998. It is hoped that the report will give librarians practical advice.
While many of the NCLIS efforts involve public libraries, there have been surveys that involve the broader community, including federal and research libraries. The rising price of journals, particularly in S&T, is being studied to inform the discussion on what the federal role in this might be, because a significant proportion of the research costs have been paid by the federal government. A Task Force has been created that is chaired by NCLIS commissioner, Jose Marie Griffiths (University Of Michigan). This TF has four tasks: 1) understanding the problem better, 2) dealing Aone on one" with the stakeholders including publishers, scholarly societies, libraries, etc., 3) bringing communities together, and 4) publishing the report.
NCLIS data is used in telecommunications legislation but they have little resident expertise. NCLIS also plays a role in international library activities. On behalf of the State Department, they administer funds which had previously gone to UNESCO. This amounts to about $100,000 per year. They have an international mission but no resources for staff. Woody Horton takes the lead as a consultant.
Discussion
NLM is trying to reach out to public libraries and questioned the definition of a public library. The number of libraries identified by NCLIS (16,000 outlets, including branches) did not match the number NLM was using. Mr. Willard indicated that NCES (National Center for Educational Statistics) establishes guidelines for what a public library is and provides the statistics NCLIS uses (http://www.nces.ed.gov). The actual number of library organizations is about 8,000, with the remaining 8,000 being branch libraries. NCES does a complete census of all types of libraries.
Mr. Willard considers the survey function to be very important to the mission of informing policy makers. John Lorenz was the statistics coordinator at NCLIS and a search is underway for his replacement. NCLIS is also involved with NCES surveys through the Department of Education.
Federal Aid to Libraries
Federal aid to libraries dates back to the Library Services and Technology Act (LSTA). This is administered by the Institute for Museum and Library Services (IMLS). Most of the money for libraries is provided through state block grants, but a certain proportion is held back and awarded to libraries who engage in leadership-type programs in research, education, or library/museum collaboration. The first round of awards ($6.5M) is being announced on October 5.
Access to Government Information
NCLIS has been an interested observer in the S. 2288 (Government Publications Reform Act) discussions. Jeanne Simon (Chair) testified early in the hearings, and generally, there is support for the legislation among the commissioners. NCLIS acknowledges that there are some operational issues that need to be worked on during the first year or two. The bill specifically requires reports on implementation problems during this period. However, the bill is not where NCLIS has put its efforts. Rather, its work with the Government Printing Office (GPO) has concentrated on the survey of electronic publications.
Back in 1994, there was a misperception that moving the government from paper to electronic would cost less. However, transitions always cost more, and people always want enhanced functionality which also costs more. GPO requested funds from Congress and directed NCLIS to do a study on electronics in government information. This was to be a rigorous analysis of the technical situation, but the money was never provided. However, Congress subsequently relented and authorized an interagency agreement between NCLIS and GPO which made this survey possible.
The survey, managed by Woody Horton (consultant) with Westat as the prime contractor, is looking at the formats and mediums used to move from paper to a more electronic environment. There is a limited budget so it is impossible to do a scientifically valid sample of the estimated 140,000 publications of the government. However, they have selected agencies and publications that they believe can inform the debate. This survey provides a snap shot of what is happening now.
The effort began with meetings with the CIOs, librarians, etc. About 400 unique products were identified from 24 agencies in all three branches of government. They also found that it was helpful to let agencies suggest publications for the list. OMB has issued a letter from the Office of Information and Regulatory Affairs (OIRA) encouraging participation.
To date, there have been two agency group meetings and a series of one-on-one meetings. The surveys are being completed. (The survey instrument is available for review on the NCLIS web site (http://www.nclis.gov/news/nclisqux.pdf) .) One survey form is expected to be completed per product. The results will be captured in a database so that others can use the information for further analysis after the report is completed. NCLIS hopes that the information gathered will serve as a way for agencies to "get a leg up" on going electronic by finding similar agencies who have gone through the process. The surveys should be returned soon; the final report is due in late February 1999.
Underlying the survey is a concern for the impact of the move to electronic products on the Depository Library System. How can the results be harmonized in order to minimize the public impact for use? The next phase will include policy implications (issues of standards versus innovations).
NCLIS Organization
NCLIS is made up of 14 commissioners who are appointed for five-year, rotating terms and two permanent, ex-officio members. The members currently include a publisher, a county comptroller, a CIO, a retired school librarian, the library director for a multi-county library system, a university administrator, a business executive from an Internet provider, a movie star and literacy advocate, a public library director, and representatives from Library of Congress and the Institute for Museum and Library Services There are eight staff including four professionals. Judy Russell has just joined Mr. Willard's staff as Deputy. He would like to improve staff expertise in telecommunications and copyright.
Discussion
Mr. Molholm asked if Mr. Willard knows of any study on the number of "fugitive documents", i.e., documents published by the agencies that have not found their way into the Federal Depository system. Mr. Willard indicated that he does not know of any. Another area of interest is the degree to which full text is available for government documents versus citations. If CENDI were interested in such surveys and would like to cooperate, NCLIS has the authority to use interagency agreements or agency personnel details.
BROADSWORD: Unifying Database Searching through a Universal Log-on
and Multi-database Search Browser
Capt. Kelly Gossen, USAF, DIA/JIVA
BROADSWORD is a multi-database search browser and universal log-on mechanism developed at the DIA Office of Joint Intelligence Virtual Architecture (JIVA), that was established to develop solutions to stimulate collaboration within the intelligence community. The technical work is being done through Rome Air Force Base. Specifically, the single interface and universal log-in was intended to solve the problem of the traditional analyst B having to have different workstations (or at least different log-on sessions) for different resources.
Built on a client/server architecture, BROADSWORD acts as middleware between the interfaces and the various databases and applications. A series of plug-ins allows the analyst to mix and match resources as needed. The databases can be unstructured or highly structured, or a mixture of homegrown and commercial database software. Common fields must be identified in order to cross-walk between the systems at a field-specific level. The interface and communication are web based. The user can also set up profiles to run in the background (as material is added to the various databases) and then e-mail is used to notify the analyst of the output.
Currently, the tool will not search across security classification levels. However, this feature is planned for the future so that users with high clearance can automatically have access to the items with lesser clearance restrictions through the same search.
The user can customize the work environment in a number of ways including setting default databases, display record formats, order of preference for a search, etc. The user can order a hardcopy or request that the item be loaded from offline to online. The results of the search can be post-processed by a variety of other utilities including PathFinder.
Even though BROADSWORD was written for a specific purpose within the intelligence community, the code is general enough that it can be modified to meet other applications for multi-database searching and universal log-on. The system administrator must configure a resource to match BROADSWORD (i.e., provide the crosswalk). It was noted that BROADSWORD could link CENDI databases. It would require an API from the interface. There must be agreement in both the access to the store-front and to the material returned. The system currently runs on a UNIX SPARC 20 or equivalent.
Discussion
Mr. James enquired about the "need to know" issue. Capt. Gossen indicated that this was a problem, but that recently a waiver was granted so that everyone who has access to Intelink is automatically considered as having a "need to know".
Ms. Hodge asked if there were any plans to add a GIS and a Gazetteer to the interface. Capt. Gossen indicated that there are no specific plans, but that these types of resources are being developed by another group.
The issue of overhead in the crosswalk searching was discussed. Capt. Gossen indicated that they have tested the BROADSWORD search versus the native searching and have found a minimal 10- to 20-second overhead.
Building the ICSTI Biological Services Enterprise through a Vocabulary-based
Gateway
Gretchen Schlag, Rick Thoroughgood, DTIC; Gail Hodge, USGS/BRD
The goal of this ICSTI project is to make international scientific and technical information more readily accessible to a broad range of users. The tool provides high level subject access to multiple databases in the life sciences using the Web (http://www.erols.com/rtgood/icsti/). Secondly, it promotes the visibility of ICSTI and its members and collaboration among the participants. The initiative was originally proposed by VINITI (Russia) and supported by BIOSIS UK and USGS/BRD at the June 1997 General Assembly. ICSTI formally adopted the project at the May 1998 General Assembly. The project is chaired by Joan Thorne, BIOSIS UK. There are currently seven participants B Zoological Record (BIOSIS UK), Elsevier (Embase), European Patent Office (EPO), VINITI (Russia), Institut de L'Information Scientifique et Technique (INIST - France), Defense Technical Information Center (DTIC - USA), and U.S. Geological Survey/Biological Resources Division (USGS/BRD - USA).
The project began with a review of the vocabularies of the participating organizations. It was determined that an integrated vocabulary was not feasible. However, a broad "topical" approach into the varying vocabularies would be possible. In February 1998, the team developed the initial two-level categorization scheme and the basic web concept. These concepts were revised at a meeting in Reston in June 1998. A prototype web site and an input form were developed by DTIC and EPO between May and July. The pilot, based on the Environment category, was developed in August. The team hopes to complete the initial content for Biology, Chemistry and Environment categories by November. The site will be registered through Yahoo, Alta Vista and the Scout Report.
The site is organized around several basic principles. There are two levels of concept description -- a high level subject category (Environment, Biology, Chemistry, Medicine), and then a short list of 10-20 keywords under each category. The web site provides a matrix of database providers to category, which identifies those providers with databases in that category.
From the matrix, the user is taken to the top keyword page for the selected provider. There is a general page that describes how the provider handles the category in general, and links to individual pages for each keyword under the category. (The frame on the side of the page allows for quick access to the keyword pages of other providers.) At the keyword level, the provider may choose to include scope notes, entry terms to the provider's vocabulary, suggested search terms, or search strategies. The terms may be linked to the provider's Web-based vocabulary or database, if the provider desires.
This matrix approach gives the provider autonomy to treat the topic as he/she sees fit within a general framework, and to link to whatever resource best fits his or her business model. The commercial providers will likely link to a commercial site, where billable transactions and subscriptions may be necessary. If the thesaurus is copyrighted, the provider may choose to bring back only a "piece" of the thesaurus or only certain elements from it. The U.S. government sites will provide free access to their resources.
A demonstration showed the treatments of the Environmental topics for DTIC and USGS/BRD. The USGS/BRD link is direct (through search scripts) to the CERES/BRD Biodiversity Vocabulary, which is currently under construction. The DTIC keyword page is linked to the public STINET database.
Several lessons have been learned from this project. Even developing a few categories is difficult, especially when crossing over specific disciplines (dealing with something as broad as the life sciences) and across multiple languages. The audience is also a concern along with the degree of jargon that should be used. The inclusion of common terminology for more scientific terms is being handled by providing common terms as synonyms on the keyword page. Different links between the vocabularies and databases are necessary in order to satisfy differing intellectual property concerns and business models.
On the whole, this project has proven very successful, and it is hoped that it will serve as a model for similar projects in other disciplines.
FROGWEB: Building an Internet-based Rapid Prototype for a Specialized User
Community
Karen Kaye, Program Manager, USGS/BRD
This project, jointly sponsored by the USGS/BRD and the Wildlife Federation, provides information on amphibians for a variety of user groups. There are a number of contributing agencies, all of whom are interested in amphibian decline because they serve as an indicator species -- an indicator that something is wrong in the environment. The site includes scientific information, including monitoring data from the North American Amphibian Monitoring Program and the North American Reporting Center for Amphibian Malformation. The children's educational information, including the Frog Quiz developed by one of the scientists at Patuxent Wildlife Research Center in Laurel, Maryland, has been particularly well received. Pictures are a key component of the site, since they can be printed out and used by the public to identify the species observed. It is important to educate the public in this area, since often the occurrences of amphibian malformations are first discovered by the public rather than by scientists.