Project to make Internet searches more credible
A new project at the University of Washington in collaboration with the School of Information Studies at Syracuse University is aimed at addressing what is perhaps the most difficult problem in evaluating information gathered on the Internet: credibility.
Michael Eisenberg, professor and dean emeritus at the UW Information School, will be working with R. David Lankes, associate professor at SU's School of Information Studies and director of the Information Institute at Syracuse, to create what they are calling the Credibility Commons. The Commons is envisioned as a place where researchers can come together to develop ideas for improving the credibility of Internet searches and Web information. It is also envisioned as a location for making different Web search tools and collections available to the public to raise awareness regarding issues of credibility of the information that is retrieved.
"One question we intend to address is whether there are ways to automatically sift and sort information to help people who are performing searches to gather the most credible information," Eisenberg says. "Most popular searches that people use simply rank sites based on how many Internet links they have in other words, how popular they are. This tells the searcher nothing about whether the information there is authoritative, documented, complete or unbiased."
The UW Information School hosted a conference a year ago, sponsored by the American Library Association's Office of Information Technology Policy with funding from the MacArthur Foundation, in which credibility of Internet information was the major theme. This is particularly a problem for students, who rely increasingly on the Internet for most if not all of their information when they are writing papers. A study conducted by researchers at Colorado State University revealed that, just 27 percent of 500 Web sites used by Colorado high school students to conduct research were judged as reliable.
This is not just an academic issue, Eisenberg says. "We know that more than 80 percent of high school students turn to the Internet first when they do research. People use the popular search engines because they are easy to use. Our goal is to develop tools that are equally easy, but also establish criteria for returning information that is highly credible."
As examples of highly credible information, Eisenberg cited the digital reference services developed by libraries around the country, the Internet Public Library, and the Ask-a-Mathematician site, maintained by the Millennium Mathematics Project.
"If we mine those sites that librarians and professionals in a variety of fields use, they might outperform conventional searches in terms of credibility," he says.
Eisenberg and Lankes intend to collaborate with a variety of experts in the information field, including those conducting research at private corporations. "There are a lot of people interested in issues of Internet credibility, but to date there hasn't been a lot of collaboration. We're out to change that."
Eisenberg acknowledges that some commercial enterprises will balk at measures of credibility. "Some sites offer paid listings for prominent placement in searches," he says. "And some of those sites do not provide clear information to users on which listings have been purchased and which ones haven't. Providing authoritative information works against the trend of blurring the sites."
"As citizens of an academic community and of the Puget Sound region, we know computer users have free access to a tremendous variety of highly credible journals and other authoritative information," Eisenberg says. "Right now, many of those sources are somewhat more difficult to find than using your average Web search. We want to make those sources as accessible to the people in our community as any Web search is now."
The current project is funded by the John D. and Catherine T. MacArthur Foundation for two years, but Eisenberg and his colleagues will be seeking additional funding for some of their longer-range goals.