There is an expectation among information seekers that all content should be available instantly via the Web, according to Mary Woodley in “Crosswalks, Metadata Harvesting, Federated Searching, Meta-searching: Using Metadata to Connect Users and Information.” While not easily achieved, meeting this demand is a concern of information professionals. As discussed in previous blog posts, the way to connect information resources with information seekers in a digital environment is made possible in part by robust metadata.
Interoperability is a term frequently heard in information studies. Woodley refers to the DCMI Glossary definition of interoperability: the ability of different types of computers, networks, operating systems, and applications to work together for information exchange. This is what is needed in order for seekers of information to find content on the Web.
An aspect of metadata creation that restricts interoperability is the tendency among cultural communities to use localized metadata structure. This makes sense, after all, metadata can be most useful when it accurately describes a particular resource using the same vocabulary as the community patrons. However, when digital resources are made available on a broader scale such as the Web, these esoteric descriptive terms can be problematic. In the Getty Introduction to Metadata, Mary Woodley describes attempts at reconciling the community metadata schemas with broader availability by use of crosswalks, metadata harvesting, and federated searching.
Crosswalks are a comparison of metadata elements set in table format. Crosswalks are the visual and textual illustration of the mapping process. Mapping is comparing / analyzing two or more metadata schemas. Crosswalks can be a great tool for converting metadata created by one standard such as VRA Core into a more general metadata standard such as Dublin Core; however, there are problems with using crosswalks as pointed out by Woodley.
Problems in converting metadata schemas via crosswalks:
- inaccurate matches
- indistinguishable original work and surrogate record
- data in one element maps to several elements in new schema
- and vice versa; data in several elements maps to one
- data forced into unrelated element due to no equivalent
- original schema employs a mix of standards
- differences in granularity
- differences in structure: some are flat, some are hierarchical
Another tool for aiding interoperability is metadata harvesting. This involves the gathering of metadata records from various institutions, databases, and repositories. These records are then aggregated into a single data base with links back to the original site. An example of this is the OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting).
Also known as broadcast searching, parallel searching, search portals, and federated searching, this involves maintaining a community metadata element sets and offering a search interface that searches across heterogeneous databases. The NISO definition of meta-searching: search and retrieval to span multiple databases, sources, platforms, protocols and vendors at one time.
I’ve found the Getty Introduction to Metadata to be very clear and instructive in its descriptions of the functions of metadata. This portion was no exception.
The Getty Introduction to Metadata
“Crosswalks, Metadata Harvesting, Federated Searching, Meta-searching: Using Metadata to Connect Users and Information”
by Mary S. Woodley