Cataloging Futures

Google Books metadata redux

The Google Books metadata controversy resurfaced recently in a Salon article by Laura Miller, The trouble with Google Books.

I focused a lot on this issue last year. It's fascinating to watch non-librarians, in this case UC-Berkeley professor Geoffrey Nunberg, arguing and explaining the importance of bibliographic metadata. Every cataloger/metadata librarian who wants to feel good about what they do should read this article.

One of my takeaways from the Google Books metadata "mess" is that full-text searching is not a substitute for accurate metadata. If it was, Google would not be spending any time or energy creating or providing metadata for Google Book Search.

It seems we need both full-text searching and metadata working in unison for optimal discovery of digital resources.

More on Google metadata

I'm convinced that the future of library cataloging and metadata creation will look something like what Google is doing now with their metadata for Google Books.

So, I was keenly interested in Eric Hellman's new post: Google Exposes Book Metadata Privates at ALA Forum. This is the most detailed description of Google's metadata workflow to date.

Why we still need classification in the age of Google

Life has trumped blogging for me lately. Just thought you might be interested in what's at the top of my reading pile: a new post on the VocabControl blog titled "Google is not perfect."

Perhaps I am starting to suffer from “deformation professionelle”, but I am constantly surprised by how often I am still asked “Why do we need classification now we have free text search and Google?”. This post is designed to answer the question. If you are an info pro, it won’t tell you anything you don’t already know, but as always I’d appreciate suggestions and additions.

The question seems to me a bit like asking “Why do we need scalpels now we have invented scissors?”. Scissors are a brilliant invention and they do many wonderful things - just like Google - they make all sorts of cutting quick and easy, but there are also many situations when they are not the right tool for the job. I don’t want a surgeon cutting me open with scissors except in a real emergency.

Fly and hope, or, What about interlibrary loan?

Matthew Beacom comments on Sergey Brin's recent NY Times editorial concerning the Google Books deal:

"Today, if you want to access a typical out-of-print book, you have only one choice — fly to one of a handful of leading libraries in the country and hope to find it in the stacks."

What the hell is he talking about? There is no need to fly and hope. He must know that you have at least one other choice: use your computer to 1. look up the book in WorldCat to see what libraries have copies 2. email your local library to use its inter-library loan service to get the book for you. He can't be ignorant of this--Google has a deal with OCLC that joins Google Book Search and WorldCat--so why did he say "fly" and "hope"? One effect of this: it makes me wonder if the other things he says are just as fishy as this. I don't know anything about those other things, but seeing what he said about the one piece I do know about makes me doubt everything else he says.

Google, metadata, and our "central task"

In light of the Google metadata controversy, Mark Pesce's post from last year is worth reading:

The central task of the librarian – if I can be so bold as to state something categorically – is to bring order to chaos. The librarian takes a raw pile of information and makes it useful. How that happens differs from situation to situation, but all of it falls under the rubric of library science. At its most visible, the book cataloging systems used in all libraries represents the librarian’s best efforts to keep an overwhelming amount of information well-managed and well-ordered. A good cataloging system makes a library easy to use, whatever its size, however many volumes are available through its stacks.

It’s interesting to note that books.google.com uses Google’s text search-based interface. Based on my own investigations, you can’t type in a Library of Congress catalog number and get a list of books under that subject area. Google seems to have abandoned – or ignored – library science in its own book project. I can’t tell you why this is, I can only tell you that it looks very foolish and naïve. It may be that Google’s army of PhDs do not include many library scientists. Otherwise why would you have made such a beginner’s mistake? It smells of an amateur effort from a firm which is not known for amateurism.

It’s here that we can see the shape of the future, both in the immediate and longer term. People believe that because we’ve done with the library, we’re done with library science. They could not be more wrong. In fact, because the library is universal, library science now needs to be a universal skill set, more broadly taught than at any time previous to this. We have become a data-centric culture, and are presently drowning in data. It’s difficult enough for us to keep our collections of music and movies well organized; how can we propose to deal with collections that are a hundred thousand times larger?

Via John Connell: The Blog

Google's metadata questions - they're our questions also

What I personally find so fascinating about the Google Book Search (GBS) metadata conversation is that Google's metadata team is grappling with the same issues that metadata librarians are dealing with also (if not right now, they we will be in the future).

• Where do you get your metadata from?
• At what point in the digitization workflow do you capture/create metadata?
• If you outsource metadata creation, do you do quality control in-house and how much (or do you leave it up to the outsourcing company?)
• Do you see metadata capture/creation as a one-time-only process or as an iterative process?
• Who (or in the case of automatic extraction, what) creates the metadata?
• Do you build in human review after the automatic extraction process?
• What is the users' role in metadata creation?
• If you have several sources to choose from, what's the best source?
• If you have several records to choose from, what's the best record and can you automate this choice?

Google responses to metadata "mess"

Google weighs in. Jon Orwant (manager of the Google Books metadata team) responds to the criticism. His comment is lengthy but well worth reading. It's obvious that the Google Book metadata team has quite a gargantuan task: aggregating metadata from 100 providers for over 168 million digitized books.

I found this quote interesting:

Geoff also suggests that "most of the misdatings are pretty obviously the result of an effort to automate the extraction of pub dates from the OCR'd text." However, we don't extract publication dates from OCR. Every misdating came from a human — some inside Google, most outside. Where the misdates come from the frontmatter (e.g., the frontispiece or the title page, as in the two examples Geoff cites) the error is more likely to have been a person inside Google. We are investigating the best ways to fix these — through better training for those people, through automated ways to identify the errors, and maybe someday through user-supplied metadata corrections. [emphasis added]

Metadata problems at Google Books

Sounds like Google needs to hire some good, professional catalogers.

In Google Books: A Metadata Train Wreck, Geoff Nunberg enumerates various inaccuracies in Google Books metadata. He argues that the lack of metadata quality really impedes scholars' use of these digitized books.

Google Using LC Subject Headings

One can't help see the irony here. Just a little over a year ago, the Calhoun Report [PDF] predicted that we would be doing away with Library of Congress Subject Headings (LCSH) in favor of keyword searching a la Google and Amazon. Now it appears that Google is using LC subject headings to enhance the search capabilities of Google Book Search. Interesting. Jenn Riley takes a close look at this new feature over on her Inquiring Librarian blog.

Scanning, Metadata, and the Google Books Project

You'll want to read this article, Inheritance and loss? A brief survey of Google Books, by Paul Duguid. It illustrates what happens when the nature of books and their accompanying metadata are ignored when scanning and digitizing.

Duguid's main argument focuses on the quality of the scanning. But he also demonstrates how the lack of sufficient metadata--something as basic as volume numbers for multivolume works--can cause problems when searching. This gets to the heart of the "digital fundamentalist" argument that search algorithms will eliminate the necessity for structured metadata. As my source mentions it's "an eye-opening read."

Via Bradley P. Allen