NTIS reconsiders fed search engine

In the wake of questions about whether it would improperly charge for government information, the Commerce Department last week decided to rethink the release of a new search engine that allows users to search through millions of federal documents online.

Officials from the department and its National Technical Information Service unveiled the search engine, which is located at www.usgovsearch.com, at a news briefing May 17. NTIS, a self-supporting agency, designed the search engine through a joint venture with Northern Light Technology LLC, Cambridge, Mass., and planned to charge $30 for a monthly subscription. The engine would tap in to close to 4 million federal World Wide Web pages and other federally related electronic documents.

But the department put the search engine on hold indefinitely - the same day it was unveiled - in the wake of media reports quoting sources questioning whether a federal agency should be able to charge for access to documents created with taxpayer dollars.

Following the decision to delay the release of the engine, Northern Light decided to offer it for free until June 1, an NTIS spokeswoman said.

"NTIS is a flawed entity," said Gary Ruskin, director of the Congressional Accountability Project, which seeks to make access to government information easier. "Taxpayers ought to be able to read the documents that we [already] pay so dearly to produce. NTIS charges often high prices for those documents."

In announcing the decision to hold off releasing the search engine, Gary Bachula, acting undersecretary for technology administration at Commerce, said officials were studying whether the new product follows federal information policy.

"In announcing the product and joint venture with Northern Light, we did not anticipate any conflict with government information policies," Bachula said in a prepared statement. "But questions that have arisen...point up the need for a review of this service from a policy perspective. Specifically, press reports today raise questions of consistency with government information policy, and we want to make sure that this search service, which was to be sold by subscription, is appropriate and fits within government policy."

The new search engine connects Internet users to documents they already can find for free on the Web or to documents that can be bought from NTIS.

The search engine's developers, however, hailed the new product as a "value-added" tool for finding public information more easily, with users being able to customize searches for specific agency Web sites or for specific topics, such as health care or agriculture. "This is the kind of information that can be very difficult to find [on a general search engine]," said David Seuss, chief executive officer of Northern Light.

Ruskin said the Clinton administration and other government leaders have been slow to make it easier for the public to find federal information. For example, he said, the administration could take measures to post on the Web public laws or important congressional documents such as Congressional Research Service reports.

"They make it very hard for citizens to do their civic duties," Ruskin said. He applauded the decision of Commerce officials to revisit the release of www.

usgovsearch.com. "We think that this service ought to be free," he said. The administration "flubbed it" by deciding to charge a subscription fee, but "maybe this time they'll figure out a way to do right by the taxpayers," he said.

If Commerce officials do allow subscription fees, there is no guarantee that many people will be willing to pay to use the service, observers said.

"I'm surprised that they can find the market for it," said James Allan, assistant professor of computer science at the University of Massachusetts, which hosts a federal information search engine called GovBot. Allan said most successful search engine companies, such as Yahoo!, use advertising - not subscriptions - to make money.

Moreover, Allan predicted that the market segment that would be willing to pay to search only government documents would be small. He said GovBot has peaked at about 10,000 hits per day.

GovBot searches more than 1 million U.S. government and military Web pages, but the engine does not allow a user to customize searches or view results by topic areas or by agency. Instead, the site returns one list of Web addresses that fit the user's query. Users then have to sift through the Web addresses to find the sites that interest them most.

Allan said the university has no plans to improve GovBot soon.

"It might happen, [but] it's not a high priority," he said. "It's something we did sort of out of curiosity." Allan said the university has sought to spin GovBot out to one private company, but no deal has materialized.