I just finished Murtha Baca’s Introduction to Metadata, which oddly enough I had read before and then promptly forgotten. Re-reading was thus a review of things I already knew, but I still feel it was useful. Pretty dry reading, but it got my head in the right place, was an excellent refresher, and provided me with an excellent reading list with which to go forth.
The most interesting chapter of the book is Mary S. Woodley’s Crosswalks, Metadata Harvesting, Federated Searching, Metasearching: Using Metadata to Connect Users and Information. At my current job we’re currently looking at demos of federated search and discovery products. Harvesting metadata vs. indexing content (having an in-house index of metadata) always comes up in these product reviews. Everyone thinks they have the right way of doing things.
The problems inherent with Z39.50, as discussed on pages 51-52 of Baca, are unfortunately exacerbated (from a user standpoint) by these in-house indexes built by organizations like Ebsco and OCLC. While the initial search (of the indexed content) is fast, and the results reasonably precise and ranked by relevancy, it only really satisfies if all the library’s holdings are in the index. Anything that has to be pulled in via Z39.50 shows up in the results list much more slowly, and depending on how the product designers handle the incoming results, can be very confusing for the end users. Libraries pay good money to subscribe to a wide variety of research databases, they don’t want a federated search product that’s only pulling in half the content that they are paying for, or that makes indexed content more accessible than non-indexed content.
Unfortunately, what I did not get from this chapter was any sense of a solution. Crosswalking has its own inherent problems, and proprietary information companies (each with their own patented technologies) make the idea of true interoperability even more problematic. Are we looking to achieve interoperability just within the world of information repositories? This would only be possible if every repository (library, archive, museum) actually owned their content and had the resources to attach metadata to it. But, at least in the case of libraries, much of what we provide our patrons digitally is merely leased or subscribed. Interoperability on the web, or on the hidden web will have to be a topic for another blog post.
Recent Comments