Friday, April 6, 2012

Re: [ACAT] Death of the "Dictionary Catalog"? Was: Name order in access points

Posting to Autocat

On 05/04/2012 22:50, Goldfarb, Kathie wrote:
<snip>
First, I want to say keyword searches can be wonderful. I just don't want to throw out the baby with the bathwater.

I want to provide a catalog that offers both options. I want to put in a word for one case when the dictionary catalog is very useful. When searching for information ABOUT an author and his/her works, the use of the subdivision can be helpful, e.g. a search for Poe, Edgar Allan brings up:
Poe, Edgar Allan -- Aesthetics 
-- Appreciation 
--Chronology 
-- Criticism and interpretation 
And many more subdivisions. Who would have thought of searching the keywords edgar allan poe influence? And, if you are doing a keyword search in a large database, there will be a lot of entries, not only those by and about Edgar Allan Poe, but many other records were some or all of those words appear, anywhere in the record.

I personally don't care if it is first name first or last name first, though that dictionary list may help them find the 'right' author, despite misspelling his/her name.

A keyword search for edgar allan poe in WorldCat brings up 21,093 records. A Subject search: 6,984. We use WorldCat Local (the free version), and I find the options to limit that search to be, well, limited. I don't doubt that many of our students would find titles they can use. I am concerned about them not finding more titles they can use because they are too far down the list. I am concerned (we are a Community College, Freshmen and Sophomores) that our students are not that good at doing the complex searching they may need. Weighting of the search terms in a keyword search can be crucial.

Just some comments from one of the old fogies that got her initiation filing in a card catalog.
</snip>

I completely agree that this is very powerful. This traditional type of access provides a truly "conceptual" search and has nothing to do with "text searching" since, as you point out, "Who would have thought of searching the keywords edgar allan poe influence?" It also provides a kind of access that is found nowhere else on the web.

But it doesn't work in today's keyword environment. When I have tried to demonstrate this to people, it is exceedingly difficult for them to understand it and I don't think it's just my own problems as a teacher. If and when they do understand, they genuinely like it and begin to question what keyword results with so-called "relevance" ranking really are. But they still do not like to search with a left-anchored text string search and will not do it. People who do not get this are not stupid, but I think it represents a fundamental disconnect between the way people think today, and the tools we make. What we need to do is figure out how to transfer this power from the library catalog into the modern world people now inhabit. Alphabetical order is one method that has mostly fallen away, but there are other options today and we should try to take advantage of them. Here is something I did, taking the subject subdivisions of Poe from the LC catalog, put them into Wordle http://www.wordle.net played a bit with it and made them into a word cloud. Here's what I got: http://www.jweinheimer.net/images/poesubjects.png.

Not perfect, but not bad for just a couple of minutes' work and a very first attempt to provide a searcher with at least some kind of idea of what is available without having to know the words to search beforehand, or "conceptual access": Homes and haunts, Raven Parodies, Knowledge Occultism, Alcohol use Physiological aspects, Ethics, and so on. This type of access actually spurs the imagination. In action of course, all of these terms would be clickable and do the correct search for you. I think this is certainly better than a regular keyword result.

If you go to Worldcat Identities for Poe http://www.worldcat.org/identities/lccn-n79-29745 and scroll to the very bottom, there is something similar but still quite different. Nevertheless, all of this can be improved immensely and could be one possibility toward taking a single aspect of the power in the catalog and transferring it into the modern world of information to provide something found nowhere else. And thereby perhaps make library cataloging a bit more important and respected?

No comments:

Post a Comment