Open Access News

News from the open access movement


Saturday, April 16, 2005

More on Google Scholar and link resolvers

Norman Oder, Google Scholar Links with Libs, Library Journal, April 15, 2005. Excerpt: 'Google Scholar may be a useful tool, but when users search for an item, how do they know it's in their library's collection? When Google Scholar launched last year..., librarians working with the OpenURL standard—which helps resolve links in a user-specific fashion—recognized that link resolvers and Google Scholar would be a good match. Since February, some 28 libraries, mainly at U.S. universities, have been testing institutional access in a pilot project, using the link resolver product each has purchased. If a user is working at a computer in the library, the access information comes up automatically; if not, the user must set specific preferences when using Google Scholar. I don't think it'll compete with traditional A&I databases that are done well, but I imagine it would definitely compete with any providers that are searching the open web," observed John McDonald, acquisitions librarian at Caltech....Once a library authorizes the company providing the link resolver to give Google its holdings, Google then highlights links in Google Scholar results pages that cite items the library holds....While journal articles that have PubMed IDs and DOIs (digital object identifiers) typically turn up appropriate library links in a Google Scholar search, neither Acharya nor librarians interviewed could estimate what percentage of results lack those links. "Items like preprints and anything from institutional repositories are indexed in Google Scholar, but they're not going to have those identifiers," Caltech's McDonald said. "Libraries are going to have to work with Google to make sure there are other identifiers." '

(PS: On the last point, let's distinguish labels from context-sensitive links for link-resolvers. Labels would help. Searchers would like to know which articles coming up in a search are OA. But the OA content in repositories won't need context-sensitive links like the priced content. That's the whole point of OA. All users have access already and needn't be steered to a copy their library has bought or licensed.)