Too Late or Just Right? OCLC, I-Schools Announce Reference Extract Web Search Project

Norman Oder -- Library Journal, 11/10/2008

$100,000 planning grant
Results would be weighted to sites used by librarians
It will draw on digital reference; can it scale?

Go back to the
Academic Newswire
for more stories

OCLC and the information schools of Syracuse University and the University of Washington have announced a Reference Extract n “effort to explore the creation of a more credible Web search experience” based on librarian input—an effort that might fill a need but also raises questions about timing and scalability. A finished project is obviously some time away, but the partners have a $100,000 grant from the John D. and Catherine T. MacArthur Foundation to fund the planning phase of the Reference Extract project.

“The best search engines are great for basic search, but sometimes the Web site results lack credibility in terms of trust, accuracy and reliability,” said Mike Eisenberg, Dean Emeritus and Professor at the Information School of the University of Washington and a lead on the project, in a news release. “So, who can help? Librarians. If a librarian recommends a Web site, you can be pretty sure that it’s credible. RefEx will take hundreds of thousands of librarian recommendations and use them in a full-scale search engine.” The recommendations would be drawn from digital reference work.

Users of Reference Extract would enter a search term and receive results weighted toward free sites most often used by librarians at thousands of institutions. "RefEx at this point 'only' covers the free web," Eisenberg told LJ. "But, there is no reason that down the road, library databases could not be included."

R. David Lankes, Director of the Information Institute of Syracuse and Associate Professor at Syracuse University’s School of Information Studies, told LJ, "The initial focus is on free web, but the point is to build a technical infrastructure and partnership model so it can cover library referenced resources regardless of type." He noted that using resources chosen by librarians would make it "very different than just dumping database contents and collections into a search engine and hoping for the best."

Scaling up
“The only way this will work is by making a project of an entire community,” Lankes said. Such efforts in the library world, however, began more than a decade ago and never completely scaled. At the project web site, one commenter observed that using del.icio.us for some searches can return “results that are far more usable than many Google searches.” The commenter questioned whether the project would scale, and also asked how RefEx would “avoid the problems that Yahoo! originally had when it planned to manually catalogue web sites?”

Still, as Google has become the world’s default search engine, various efforts have emerged to serve smaller subsets of users with more credible information. Recently, as LJ reported, the semantic search engine Hakia has made an open call for librarians to help bolster its “credible sites” results section. The company is offering prizes; each “credible site” submission makes librarians eligible for a raffle of $500 worth of books donated to their library, or one of two $1000 conference grants for submitting the most URLs.

Would lures be needed for Reference Extract? No, said Eisenberg. "What’s unique about our approach is that it doesn’t require additional effort (or carrots to get librarians) to contribute. We are automatically gathering URLs from their digital reference work."