Every time I look, metasearch is still with us. Part of me keeps hoping it will go away, but nope, it’s still there. And thank goodness that there are enough people and companies out there still trying to make it better.
Index Data announced Monday that they have created IRSpy, a registry of information retrieval targets that support Z39.50 and SRW/SRU. “Each registry entry consists of both database level information, such as the indexes supported for searching and the record syntax and schemas supported for retrieval, and record level metadata such as titles, authors, and descriptions.” Such a registry has long been needed by the metasearch industry. It could even fulfil another need—the creation of a list of targets that adhere to standards, thus distinguishing themselves from those that still require screen-scraping, HTML parsing, and other technical hoops that continue to support a metasearch middleware industry.
Regardless, the middleware industry continues, and continues to improve its software. Most vendors have implemented some sort of clustered or faceted browsing into their interfaces. Those that haven’t have plans in the works: WebFeat announced several new product offerings at ALA this year, including more robust administrative tools, a proxy server for single sign-on, and an upgrade (version 3.0) to WebFeat Express.
Not to be outdone by the open source ILS industry, two projects continue to get traction in the open metasearch environment. IRSpy, mentioned above, is a nice complement to Index Data’s re-engineered MasterKey. LibraryFind, which includes features like an OpenURL resolver and the ability to index local collections, is another option from the Oregon State University Libraries. LibraryFind was developed using Ruby on Rails, which has become quite popular among open source developers.
Lest we forget, there is still a large contingent of people who hope that the need for metasearch will one day be obviated. Google Scholar is still moving (though hardly hurtling) toward that goal with the recent inclusion of ScienceDirect content. Rather than repeat it, I will simply point to Peter Brantley’s nice commentary. What will become of Scopus? If we’re hurtling toward anything, it might just be one of the first viable pay-by-the-drink models for scholarly content. As if library collection budgets weren’t difficult enough.
As OCLC Openly’s Eric Hellman once commented at a NISO meeting on metasearch standards, “Metasearch will work perfectly when all the data is in one database.” Until then, we struggle onward.
[This post originally appeared as part of American Libraries’ Hectic Pace Blog and is archived here.]