The new ResearchWorks section of the OCLC Web site (http://www.oclc.org/research/researchworks) gives interested parties a chance to preview and pretest demonstrations, prototypes, and other interactive examples that showcase current explorations and developments coming from OCLC Researchers. Now visitors can explore new features and services that may—or may not—debut on the full service in time. The site encourages users to provide comments and even share those comments with others in online discussion groups. Currently twelve OCLC Research projects had active demonstrations available: bookmarklets, Extensive Repository Response Locators (ERRoL), xISBN, an "info" URI service, and LC Name Authority File. Rather than hold new developments as proprietary secrets, OCLC openly encourages observers to treat the information it provides "as examples of things you can develop or incorporate into your own systems." OCLC also warns that "these demonstrations are still in development, and they may perform erratically or disappear without notice" and suggests return visits.
All 12 demonstration projects listed on the ResearchWorks home page carry icons identifying whether the project is designed for human-to-machine or machine-to-machine use, or both. Five of the 12 carry background information ("Learn more...").
Jon Udell's LibraryLookup bookmarklet can help match titles located on online bookseller sites, such as Amazon.com or barnesandnoble.com, with holdings of local libraries, but Udell's bookmarklet only uses a single ISBN. OCLC has integrated its xISBN lookup service to locate related ISBNs and feeds the set of ISBNs to the local library catalog for checking.
The xISBN demonstration lets users enter an ISBN (http://labs.oclc.org.xisbn/[ISBN]) to retrieve a set of associated ISBNs. The machine-to-machine could help library staff locate all editions or printings of a work before placing an order or confirm that no alternative versions had already entered the collection.
The ERRoL project provides an identifier resolution service for Open Archives Institute (OAI) repositories registered with the OAI Registry at the University of Illinois at Urbana-Champaign. The URL pattern in ERRoLs incorporates OAI identifiers that users can extend and use in place of PURLs and Handles. Although based on the functionality in OCLC's OpenURL registry service, ERRoLs open the OAI "extensions" to any OAI repository with a unique "registered" repository.
The "info" URI Registry provides a way to translate library identification schemes, e.g., Dewey Decimal Classification, LC control numbers, OpenURLs, etc., into Uniform Resource Identifiers. Each URI begins with a scheme name. The registry supports the integration of information from legacy systems by Web-based applications, such as Xling, RDF, or Topic Maps.
The LC Name Authority Service expedites checking names in the Library of Congress' Authority File. An Interactive Response Type works for entering names one at a time while the Automated Response Type handles batches. Again, both aim at machine-to-machine interactions rather than human-to-machine. OCLC intends the service as a tool for institutional repository software, such as DSpace, ePrints UK, CONTENTdm, eprints.org, Fedora, etc., which need to provide consistent and knowledgeable metadata.
The remaining seven demonstrations carried brief descriptions:
Locate fiction for your reading pleasure. Search by your favorite author, title, genre, and more.
Open URL Registry (11/26/03)
Implementing OpenURLs? Register your identifiers here so others can use them in OpenURLs.
4.2 million electronic and non-electronic thesis and dissertation records extracted from OCLC WorldCat.
RDF Translation Tool (11/26/03)
Translate Dublin Core metadata vocabulary into your own language.
NDLTD Union Catalog (11/26/03)
50,000 metadata records for electronic theses and dissertations harvested from libraries around the world.
DCMI Registry (05/03)
An authoritative source of information about the Dublin Core element set and related vocabularies.
The rich vocabulary of LCSH in a simpler syntax.
Although encouraging librarians to take advantage of ideas demonstrated in ResearchWorks, entry to the section requires agreement with a lengthy set of terms and conditions. Beyond the usual "ride at your own risk" legalities, the terms maintain ownership of the content and software as the property of OCLC and friends. Visitors may "not copy, reproduce, alter, modify, create derivative works, or publicly display any content (except for your own non-commercial use) from the Site." Nor may they "use any robot, spider, other automated device, or manual process to monitor or copy any content from the Site." On the other hand, OCLC may use any comments or suggestions from visitors to improve or change its products without offering any "compensation, partnership, or recognition beyond the association of the comment with its author in the forum in which it is offered."
Not everything users see in ResearchWorks will graduate into an OCLC production service, though the company does admit that wild response to a beta test would push it toward integrating it with the main service. Comments go directly to the OCLC scientist working on the project.
If the new OCLC ResearchWorks looks familiar to Google aficionados, it should. The FAQs for ResearchWorks openly acknowledge Google Lab's contribution:
We were inspired by Google to mount these demonstrations of OCLC Research's works in progress. Google graciously gave us permission to use their Labs page, including their terms and conditions, as a basis for ours.