Quest, Present and Future
Since the beginning of the twentieth century, library catalogs have been constructed according to a set of principles and techniques that has remained largely unchanged. From the 3 x 5 card stored in wooden catalog card drawers through the early automated catalogs and on to the current web-based OPACs (Online Public Access Catalogs), a catalog record provides descriptive data and a series of access points, or browsable index entries, under author names, titles, series, and subjects. Keyword searching and limiters such as date, format, and location have given users added techniques to locate needed materials. More recently, the ability to link directly to an electronic resource has enabled the catalog to deliver the full text of a resource to the user’s desktop.
And yet users prefer to search for information resources on the Web rather than in our OPACs. Recent and dramatic changes in our collections, in technology, and in user expectations have encouraged discussion on the role and scope of the OPAC. Our library catalogs have always described our collections as containers or bundles of content: books, journal titles, physical media. Reference collections held printed indexes to periodical articles and individual works of poetry, book reviews, essays, etc. But the collection is no longer in the library. Physical collections are stored offsite; scholarly materials are in repositories and open archives; authoritative resources are scattered on the Web; e-journals and databases are on vendor and publisher websites. The OPAC no longer represents the full range of scholarly and research information available to the university community.
Library automated systems were designed initially for inventory management: acquisitions, cataloging, and circulation. The MARC (MAchine Readable Cataloging) format was developed in the 1960’s to support the exchange of catalog records and the printing of catalog cards. While MARC has proved extremely durable and adaptable, other encoding standards, or metadata schema, have developed outside the library to support the electronic exchange of information. These schema are both simpler and more detailed than MARC, serving the specific needs of their communities: the publishing industry, museums and archives, and specialized collections. There is much research and development on “crosswalks,” or conversion routines, to map one metadata schema to another (including MARC). But given the fragmentation of our collections noted above, it is worth asking whether we should be converting metadata into MARC in order to include it in our OPAC, or finding ways to build on the strengths of these specialized schemes through improved search interfaces.
The public has learned to interact comfortably with the electronic world. Users have taken control of searching and exploring on their own through their preferred interface. They participate actively in creating content and sharing it with other users. The availability of full-text resources has made it possible for users to seek the most granular, or specific, level of information rather than our traditional “bundles.” For many users, the process of discovery, or search, is a means; delivery of content is the goal. But the discovery process is greatly improved when it relies on descriptive metadata to retrieve and rank results in a meaningful way.
How can we improve the user experience in our OPAC and still take advantage of the rich descriptive data and expert categorization and classification created by catalogers?
One approach is to separate the discovery process from the internal management functions by putting a search engine on top of the library system, using cataloging and other metadata to present results in new ways. Enrich the data in the catalog record by adding tables of contents, reviews, and images to maximize successful retrievals. Provide active feedback to suggest alternate searches, correct spelling, and link to the best source of full text. Cluster results in hierarchical or graphical ways that make relationships clear and suggest paths to refining the search. And link, link, link … the cataloged collection to repositories to journal articles to digital objects to other catalogs, seamlessly, and return consolidated results to the user ranked by relevance or by whatever sort order the user chooses.
These developments are already underway by a number of system vendors, including BC’s vendor, Ex Libris. No matter which discovery system is used, it relies on metadata for relevant results. The richest source of metadata is the library catalog with its bibliographic description and controlled vocabulary. The new search engine OPACs will gather metadata in MARC and other formats, consolidate the search results into a coherent view, and present options to users to assist in selecting the appropriate resources to meet their information need. And through projects like OCLC’s Open WorldCat, our cataloging metadata will be harvested by the Web search engines to meet the users on Google and Yahoo and lead them back to our collections.
Head Librarian, Cataloging Department