November 18, 2003

Stupid code

Posted by Eszter

I was searching for the journal The Information Society yesterday on our library’s online catalog system. I had looked up the journal the day before so I knew that we had a subscription to it. Regardless, yesterday I kept getting “Your search found no matching record”, which was incredibly frustrating given that I had just browsed the journal the day before. Finally, I decided to try the search without the “the” in the title. I’m not sure why that occured to me, but I gave it a try. Surprise-surprise, searching simply for information society specified as the Journal Title worked.

I realize search engines often exclude articles like “the”, but it seems if it is part of a journal title it should not count against you to include the article. It’s one thing to exclude it completely but bring up related results nonetheless, it’s another to have it count as a hindrance leading to no results. In case anyone’s wondering, using quotes or a plus sign in front of “the” - strategies that would help in some search engines - do not lead to any results either (in fact, the quotes confuse the system completely).

I wrote to the library to tell them about the incident and ask whether they could tweak the code so it wouldn’t count against you to know the actual full title of the journal one is seeking. I just received a response according to which the library has already asked the online catalog system provider Endeavor to make this change - as have other libraries apparently - but they have not cared to improve the system. Ugh.

Posted on November 18, 2003 06:46 PM UTC
Comments

You would not believe (or perhaps you would after all) how difficult it is getting catalog system providers to make even the smallest and most rational changes.

At ALA one of the primary topics of conversation, in the bars after the meetings, is just how lousy catalog system providers are. How unresponsive. How take-the-money-and-run. How illogical. How sullen and uncooperative.

Librarians generally like our jobs, but system providers are bad, bad, bad. Hates them, my precious, we do.

Posted by Jess Nevins · November 18, 2003 07:01 PM

I wanna know what you’re feeling —
there are some things you can’t hide.
I wanna know what you’re thinking —
tell me what’s on your mind!

Posted by dop · November 18, 2003 07:22 PM

I was having the same problem today with a demonstration I did of database searching in one of my history classes: an incredibly balky catalog system for one journal database kept ignoring my requests for titles when I knew there were citations there that were relevant to the title. Making a small change in the title phrase entered turned it all up—but it was definitely a face of a stupid search design.

I think this is one of the small but massively important issues in contemporary academic life, and the passivity of academics about it drives me wild. The proliferation of vendors and interfaces in electronic databases, and the sometimes extortionate pricing of some databases, is getting to the point of being absolute unacceptably dysfunctional.

We need to seriously push our institutions and libraries as a single, united negotiating entity towards promoting a universal search interface for ALL citational databases relevant to academic work. There’s no reason why Book Reviews Online, InfoTrac, EXAC, ABC-CLIO and many others ought to be as different (and sometimes badly designed) as they are. I would argue that a united bloc pushing a single highly usable standard should be prepared to use its negotiating power to completely boycott any vendor that refuses to engage in the construction of such a standard in concert with other vendors.

This would go along with academics in toto refusing to order or provision material to participants in any full-text journal service that does not adhere to “fair price” constraints and a continued commitment to providing material that they own, but that’s another issue.

Posted by Timothy Burke · November 18, 2003 08:51 PM

I’d love to have your problems.
Seriously, this stuff is incredibly common and if you haven’t run into the problem often enough to routinely make varied queries, you should count yourself lucky and be glad you learned something new the easy way (you didn’t order the journal to later find out your library already has it).

Posted by markus · November 18, 2003 09:05 PM

Timothy - I agree that some joint approach would be nice to see. It’s a bit frustrating to hear from the library that they know other libraries have complained as well, but the vendor simply refuses to improve the product.

Markus - the point isn’t whether and how often you’ve run into such a glitch, the point is that it should be fixed so we could use our energies for other things than running repeated varied queries.

Posted by eszter · November 18, 2003 10:22 PM

On a related issue, I’ve had British journals change the spelling of references to conform to British spelling. The result, of course, is that the reference is now wrong, and searching for it under that name won’t work well.

Posted by bob cox · November 18, 2003 10:25 PM

Academics, liberate yourselves from the tyranny of closed publishers and their crappy proprietary search systems! Computer scientists already have, more or less; CiteSeer indexes nearly all the computer science academic papers freely available on the web (which, these days, is nearly all newly published papers). Because CiteSeer’s an open database, you can also use a site-focused Google search.

Do your profession a favor: create a new professional norm wherein academics commonly make all their papers available in PDF format on their home pages. Then have your professional organization license the CiteSeer code. Then, someday, you too will be able to search research papers using Google. You have nothing to lose but your chains.

Posted by Armature · November 19, 2003 01:04 AM

Yeah, I was going to say, isn’t this a problem tailor-made for Google to solve? Isn’t it almost trivial, compared with 3bn web pages?

Somebody tell the biz dev folks at Google; here’s a market to clean up in…

Posted by Doug · November 19, 2003 09:39 AM

At a guess, the journals don’t let Google in; most of them restrict access to paid users.

Journals are in a similar position that the record and movie industries are in, but worse. They make their money on controlling information flows, but the information comes from a subset of their customer base. And more critically, their review process comes from their customer base.

They’re sort of like a record company which sells to artists, or a publisher who publishes fanfic for a fan base. Their customers are in a much better position to subvert the process.

Posted by Barry · November 19, 2003 12:37 PM

Barry, I think Doug’s point may be that Google could try to go into this market.. not in its publicly available form, but in proprietary mode. It’s a good idea. And Barry, you’re absolutely right that the customer base should be able to have a pretty big say in this. I’m going to ask around to see if Google has considered this market at all.

Posted by eszter · November 19, 2003 03:13 PM

We use Endeavor here. I have no idea why they do that. Other systems (for instance Innovative, and I think Sirsi) strip out the articles automatically.

Speaking as a librarian, indeed a cataloger… I didn’t even notice this before. I reflexively get rid of initial articles anyway, because I know that they aren’t of any use. I use other library catalogs all the time. Another problem is name order; if you use an author browse and search for “Mark Twain”, Voyager isn’t smart enough to figure out that the surname is “Twain” and will send you to last name “Mark”. Our students run into that all the time.

Why not go bug them in person? They’re a Chicago company.

1350 E. Touhy Ave. Suite 200 East
Des Plaines, IL USA 60018-4505

Interestingly, the Z-39.50 interface for Northwestern uses a different search style and your search would have worked on that.

Posted by Mac Thomason · November 19, 2003 03:59 PM

At UMD we often use Dialogue which make suse of command line codes to do even a basic search. It’s infuriating to say the least, especially since I often find what I need for my Library Science classes on Google.

Posted by Keith · November 19, 2003 06:59 PM

Library catalog searches are close to my heart, since one of the first significant projects I led (in 1986) was to write a catalog and circulation system for a university library. The question of an intelligent search interface and its implementation generated an interesting set of problems. However Google seems to have by and large solved it. They do have a Google Search Appliance, hardware and software to index any collection of documents - it’s what I’d buy if I had the problem to solve again.

Posted by Douglas · November 20, 2003 12:13 AM

Eszter, that’s exactly what I was wondering. And it looks like Douglas has answered aat least part of the question. Ready for the Google IPO?

Posted by Doug · November 20, 2003 08:07 AM

So I just got asked by a patron to help them find an article on our old balky ProQuest CD-ROM system. Didn’t work so we used the online. And the article she needed was listed alphabetically under The. Had a hard time finding it scanning the journal table of contents (it was a long one). I had to find it another way, as I didn’t think of looking under the inital article. Old habits die hard.

And we didn’t buy Endevor because we hated its catalog interface…

Posted by Cathy · November 21, 2003 12:59 AM
Followups

This discussion has been closed. Thanks to everyone who contributed.