Cataloging Futures explored this resounding question last week "What happens to cataloging when the majority of the library's books (and other resources) already come with metadata?"
I recently had the opportunity to consider automatic metadata extraction and embedded metadata and I feel compelled to voice some of the same skepticism about certain interoperability and consistency issues that I have with metadata creation in general. While a smart computer like IBM's Watson could take some of the subjectivity out of applying metadata, is anyone ever going to agree on a consistent format, scheme and standards?
If the entire world of objects could be cataloged only using Dublin Core, for instance, automation would be perfectly fine. Coupled with XML and some extended controlled vocabulary standards, a system could be developed that would standardize the world. But, has that ever happened yet even in analog? No.
Someone, a person, still needs to do the thinking behind the automation of creation or extraction. Decisions will be required and people will have to make these determinations. And, those people will still disagree and tinker and innovate and discuss metadata and cataloging.
I share your skeptism with automatic metadata extraction but I think it will/can be a value part of the process in the future. If you haven't read this TechWatch report by Richard Gartner I would recommend it for provide some excellent context re: metadata and the library community <http://www.jisc.ac.uk/whatwedo/services/techwatch/reports/horizonscanning/hs0801.aspx>.
ReplyDelete