Super Web tool rescues GILS

A group of federal agencies and vendors are creating a super World Wide Web crawler and search tool that would help agencies automatically index their electronic documents and make it easier for the public to find federal information.

The software, called the Advanced Search Facility, is being developed by Pilot Research Associates, Vienna, Va., and its subcontractors under a contract with the General Services Administration. Proponents view ASF as a way to rescue the moribund Government Information Locator Service by offering agencies a free, automated tool for creating GILS-compliant records of their documents.

In addition, ASF, which the group plans to release as freeware within the next couple of months, aims to make it easier for users to find what they want on the Web.

"The problem it solves is one that's not being addressed by Internet search engines," said Paul Christy, deputy director of Stat-USA, a part of the Commerce Department that publishes economic and trade information. Stat-USA developed the requirements for ASF with the Office of Management and Budget and the U.S. Geological Survey. "Internet search engines do a great job at locating a needle in a haystack, but the government is lots of haystacks full of needles," Christy said.

William Miller, the acting chief for geographic research and applications with USGS, unveiled the project at a Nov. 4 meeting of the Federal Webmasters Forum.

GILS— an electronic card catalog of government documents— was required by Congress four years ago to help the public find federal information. Most agencies ignored the mandate because they found creating records for the system too complicated and time-consuming.

"Webmasters have time for everything but describing documents," Christy said. The $500,000 project is being funded by the Government Information Technology Services Board.

Almost all the functions included in ASF "have been available in separate packages before," said Elizabeth Huber, program manager with Pilot Research, in response to questions e-mailed to the company last week. But these capabilities have not been "as highly integrated, nor as focused on the problem of creating and managing locator records."

Christy said the tools add up to a "framework" for organizing information. The software would capture author, title, version, subject and other data from existing documents and map this information to the GILS format. In addition, agencies could use ASF to point to related sites so that when users query one database, information is returned from others as well.

William Moen, a GILS expert with the University of North Texas School of Library and Information Science, said ASF is one of many projects to develop better search and retrieval tools. A similar effort is under way with Computer Interchange of Museum Information, (CIMI) a consortium of libraries and museums that includes the Smithsonian Institution.

"We're not trying to say to every community, 'This is exactly how you have to put your stuff,' " said John Perkins, executive director of the group. "You map your own rich data into these elements in a way that makes sense to you." Moen said one problem facing the CIMI project that ASF may encounter is how to capture the right information from documents that have not had their titles, authors and other key data electronically tagged before being indexed. "You see that variability of quality," in metadata records from unstructured documents, he observed.

Huber said ASF would provide "a common starting point" for commercial vendors who want to develop interoperable systems on distributed networks. Vendors of commercial products for creating, capturing, searching and distributing documents could substitute their products for any ASF component.

A beta version of ASF is available for Unix platforms from the project Web site, asf.gils.net.