I don’t want to get into the “OPAC wars” at the moment, but reading reports on various programs, I have to ask:
What’s a “known item”?
I ask that based on the assertion by a number of people that online catalogs are only good for finding known items.
My naive sense is that a known item is one particular entity (or, pace FRBR, a work which may have multiple instantiations), identifiable by a title and, usually, an author of some sort.
In which case, I’ll assert that any decent OPAC is also good for something other than known item searching that matters to quite a few library users:
“What do you have by this author/composer/musical group?”
Maybe that is a known-item question, since it presumes knowledge of the name of the creator. But I don’t think so: The assumption is that there are multiple items, and only the creator’s name is known.
[Incidentally, the question generally is “what do you have that’s currently likely to be available to me to walk out with today,” not “what exists somewhere in the bibliographic universe.” At least that’s my public-library-patron assertion. And I do regard “oh, and by the way, so-and-so also writes under these names, and you can click here to see what’s available under those names” is a very helpful additional answer, although tossing the other pseudonyms in with the initial result may not be so helpful.]
(I’ll also suggest that many OPACs are pretty decent at “more like this” searching based on hotlinked subjects or even call # browsing, but that would get into the OPAC wars, and I don’t want to go there. Yet.)
So: Do I misunderstand “known item”? Or is the claim of uselessness possibly overstated?
From http://www.db.dk/bh/Core%20Concepts%20in%20LIS/articles%20a-z/known_item_search.htm – Kind of information search, in which the user know what documents are searched for (or where at least certain data about the documents are known, such as the title or the author).
So – Hobbit, JRR Tolkien – both are known item searches.
Subject and keyword searches are not known item searches…
Hmm. So “all those titles I don’t even know exist, by an author I like” are known items?
I guess my pseudo-librarian status is showing. That may be the actual definition, but it strikes me as, um, counterintuitive. It’s really “known access point.”
[I’ll also suggest that “mailer fire moon” as a keyword search is in fact a known-item search, where the patron just isn’t sure of the exact wording of the title. But never mind…]
Walt, I don’t use individual library OPACs to find all the works by an author I like, because the missing question is “did the library/library system actually buy them all?”
I don’t use Amazon unless the author is current, either. I prefer abebooks or WorldCat.
Call-number browsing is neat, but who understands it? I didn’t get it until I was in library school. Relabel it.
And OPACs are worse than they oughta be at the “what’s here, now, today?” question too, because it’s too hard to limit by library, and impossible to limit by “not checked out.”
All true, Dorothea–but as an ordinary library user (which is what I am when I get back in book-reading mode at MVPL), I want to know what I can get now, here, and for a lot of libraries, the OPAC tells me that pretty quickly. (MVPL doesn’t have branches, which helps.)
I agree on call number browsing as a label…and, of course, if you’re *at* the library, you don’t need it. (One of the first columns I did in PACS Review was on that function of a catalog: “Just Point Me In the Right Direction.” Actually, now that I check my vita, it was the first column, since it was in the first issue of the first volume of the ejournals…)
I’m not trying to defend OPACS–although I do believe that “How can we improve online catalogs?” may be more fruitful than the rounds of denouncing them that seem to be in vogue. It was really just this one specific issue.
I exposed my ignorance as to Proper Library Terminology. David King corrected my ignorance. Ain’t the web wonderful? (“Ain’t” has been needlessly denigrated as bad English, although in this case my usage is inappropriate or deliberately faux-causual, since “Ain’t” is properly the short form of “Am not”…and there’s another discussion that I should probably stop before it begins.)
I’d agree on your (Walt’s) definition of a known item, and OPACS are good at more than that. Most people dismiss known-item searching, but in a large catalog/library it can be a challenge, especially with incomplete information.
I once did a post blaming OPACs for only doing what the Paris Principles suggested a catalog do. Today I was reading the latest STATEMENT OF INTERNATIONAL CATALGUING PRINCIPLES which goes a lot farther (at http://www.d-nb.de/standardisierung/afs/imeicc_papers.htm) (found through Catalogablog)
–Th
Good to hear from a future colleague…
I’m coming to this post a bit late, but I would include known (with proper spelling!) author or title or composer as a known item search — the opposite in my mind is having a problem or a nebulous information need and trying to use an OPAC to find a resource that might fill that need. The overlays that are in vogue don’t change the catalog or the search, but just explain the results that are presented. That seems key to me.
As for presenting “a solution exists” vs. “here is what you can have in your hand or on your screen in 5 minutes” — it pretty much depends on what the user needs. Research libraries should probably tend toward “a solution exists”. Someone (maybe Dempsey?) was recently commenting on libraries displaying best seller lists and how that leads to disappointment as the few copies the library has may have 100s of holds. Either way, the user should be able to indicate “show me what I can get in 5 minutes, 2days, 14days” (someone else suggested this, too, but can’t remember whom to cite!)
I guess we’ll have to agree to disagree. I believe that calling a search for “all items that you have that are by a given author, whether I’ve heard of them or not” a known-item search does violence to the ordinary-English definition of “known item.” Of course, there’s nothing new about Biblish and English being different languages.
To further muddy the waters. . . .
I found a nice bunch of books on CSS in my OPAC this morning by entering CSS as a keyword search. . . so on that one it worked well for unknown item, which is kind of rare. I can usually find known items when the item is exactly known–that is, when the author’s name or title are exactly right. What my OPAC is really bad at are what I might call (in neither English nor Biblish) “sorta known items.” The other day a patron wanted to get Ken Burns’ jazz series, which he’d checked out from the library a few years back. Since I wasn’t exactly sure of the official title of hte series, I put jazz ken burns in as a keyword search and got. . . nothing. That I find frustrating.
If I was being snarky or anti-OPAC, not that I would ever be either one, I’d suggest that the CSS case is one of “a distinctive known term that is likely to occur within the limited metadata of bibliographic records,” and could thus be considered a “semi-known item.”
As for the second example: Just for fun, I entered that search in the RLG Union Catalog/Eureka, which does not have a true keyword index (it’s a long story, and since RLG no longer exists, it’s not a very interesting one: suffice it to say that the keyword index is a combined author, title, subject word index–but in all other Eureka databases, it’s a true, carefully-designed, not every subfield but most of them, keyword index. I say “carefully designed” because I did the design: no false modesty here).
“jazz ken burns” yielded 20 items, of which the second is the 10-DVD “Jazz” set. Ken Burns appears as an added author entry, but as is typical of a video, there’s no main author; that may be the problem with your OPAC. I’d find that frustrating as well.
(What were the others? The companion website and 18 sound recordings. I would note that a “relevance-ranked” result would probably not show the DVDs until the 19th or 20th position, since “Ken Burns” appears well down in the record, whereas “Ken Burns Jazz” appears as a string in the titles of most of the sound recordings. Make of that what you will… On the other hand, if “relevance” is “popularity,” then the DVD set would be back at the top, since it’s by far the most widely held. Let’s see: What does the other view of an older subset of the RLG Union Catalog, that is, RedLightGreen, show? Hmm. No good way to identify the DVD set, which in any case doesn’t appear to be on the first page…)
If it’s not obvious, I’m not saying OPACs are wonderful, not subject to improvement, or whatever. As this example may note, it’s complex–and your mileage may vary.
I might regret taking a stab at this, but here goes. I think we are making some fuss over what boils down to a poor choice of words. “Known item” really means “known title, known author, or known call number.” Is “known item” oversimplified? Yes. Easier to say? Also yes.
Now let’s get “known author” out of the “good enough” category. Have you ever looked carefuly at an authority index browse for authors? Let’s ignore the whole “last name first” problem, and just think about all those added entries and see references. What a mess. You need an MLS to interpret the average author search for Shakespeare, William. I’ll take our new keyword search version of the same search any day of the week.
Andrew,
There’s a difference between showing the whole bloody authority file and doing a simple list of author names. (Seems to me that the new search you’re sending me to is an author-word search, not a keyword search. Snazzy left sidebar, absolutely–and I’m certainly not arguing that NCSU’s new system isn’t a better OPAC, because I think it is, and congratulations to you folks, even if it did take better than half a minute to return the screen–but there’s nothing terribly new about author-word searches. Sez I, having been analyst for too long for a system that has/had them.)
There are a number of author search methods that can give a “good enough” result for most users, even if it gets clunky on a handful of extremely prolific authors (well, Shakespeare wasn’t that prolific, but the universe of books related to him is very large). I’ll agree that the NCSU authority display isn’t good enough. I think that quite a few OPACs do a pretty decent job here, though.
I’ll suggest that “known item” as expanded is an attempt to shore up a dichotomy that I don’t believe exists: Known item vs. subject search. I think there are more gradations than that, and that “titles by this author” is a very common, generally reasonably well-served middle point. Maybe not well-served in all OPACs, but certainly in some, including the not-terribly-L2 one at my local library.
And, much as I hate to pick nits, I’ll pick a couple here:
On what basis are the three most relevant books for “William Shakespeare” as an author search
Henry the Fourth, Part 1
The Tempest
and Elements of Literature (heck, I never even knew Shakespeare wrote that one!)
I note that all three have 2006 publication dates. And that the next 10 or so have 2005 publication dates. But I can’t suss out any meaningful relevance here…
By contrast, the old-fashioned local public library catalog, with 478 holdings, shows them in [wait for it] alphabetical order by title, which at least has a form of meaning to English-speaking users.
Sorry, couldn’t resist: “Relevance” in bibliographic records continues to bemuse me.
[And to repeat: Yes, I like the NCSU/Endeca catalog. Yes, I think there’s loads of room for improvement in online catalogs.]