Turning its vast cataloging expertise to taming the Web, OCLC recently announced that it will open participation in its Web cataloging pilot project, the Cooperative Online Resource Catalog (CORC), to all libraries. This will apply OCLC's highly successful cooperative cataloging model to information-rich Web sites, with the prospect of a catalog of Web content on the scale of OCLC's dominant bibliographic database, the WorldCat (OCLC Online Union Catalog.)
The CORC project originally called for a closed membership of 100 libraries, each committed to an ongoing program of Web site cataloging. There are now 170 libraries worldwide participating in the project, and the CORC database has reached 200,000 records. However, according to Taylor Surface, program director for the CORC Project, OCLC has experienced great demand from libraries for broader participation in the project. The response has been so strong that OCLC decided to redefine and expand CORC into a truly cooperative library service. According to Surface, OCLC has concluded that "CORC is worth turning into a sustainable service for libraries." CORC, now still in development, will be released as a commercial product by mid-2000.
CORC applies old and new methods to the unique challenge of Web site classification and cataloging. All libraries are eligible to join and to submit cataloging—a cooperative method that has built WorldCat into the world's largest book catalog, and a reference and research tool of unsurpassed importance. Libraries will apply their own individual collection development policies to select Web sites for CORC, as they now do for entering book records into WorldCat.
OCLC has also developed new technological tools for cataloging and updating the wildly inconsistent and constantly changing Web. Templates will assist catalogers in extracting data and creating records. Libraries can catalog with MARC, with the Dublin Code (a system for using metadata to catalog electronic documents), or both; MARC and Dublin Code records for the same record will be mapped. OCLC has created automatic checking routines for Web site URLs and content, which will be run regularly as a means to keep CORC up-to-date. Libraries will also be encouraged to monitor CORC content and report changes that they discover.
In addition to bibliographic records for individual Web sites, CORC also contains member-created "Pathfinders," or collections of topically related Web links. To maintain their timeliness, Surface explains, changes in a Web site record will automatically flow through to its occurrence in any Pathfinder. There are now several hundred Pathfinders in CORC.
CORC is OCLC's latest and grandest Web cataloging project. It was started in January 1999 as a project of OCLC's Office of Research, with 50 library participants. OCLC's first such effort was InterCat, which ran from 1991 to 1996 and created a database of 50,000 records. InterCat was followed in 1996 by NetFirst, an OCLC staff-built database that now has 100,000 records.
CORC is a bold and creative venture that applies OCLC's mission to the Web. Opening participation is not only in the spirit of OCLC's cooperative foundations; it is undoubtedly a necessity if CORC is to have any real chance of keeping up with the rapidly expanding Web. The new technological tools for cataloging and updating records are also necessary; unlike a printed book, which is eternally stable, Web sites can and do evolve continually.
The proof of CORC will be in its execution. Can CORC, with very wide library participation, maintain adequate consistency and quality control? Will CORC's maintenance tools work well enough to keep it timely? (An obsolete Web site description is less useful than yesterday's newspaper.) And finally, can even the collective labor of OCLC and its members keep up with the Web's explosive growth? These are daunting chores, but no one is better able to shoulder them than OCLC.