First the need, then the tool

In his post “My ALA Baggage,” John Blyberg says:

There is a fairly severe disconnect between what the 2.0 pundits say (among whom I count myself), and what is really happening. Your library may have, for instance, a Flickr account, IM reference, a bloglines blog, delicious bookmarks, whatever. But are they truly embedded into the way your institution works? In almost every case, this approach seems like throwing seeds into the air, letting them land where they may. I think it’s time to start talking about how we arrange these components into a more suitable constellation of services. These technical elements of L2 must be aligned along our institutions’ field of influence and expertise so that the seams don’t show. Seams send the wrong message, they say we’re being disingenuous and sloppy. In effect, poorly implemented technology amounts to spamming our users and staff with “new features.”

Yes, yes, yes! I think that librarians are still struggling to see how 2.0 technologies can be worked into the tapestry of services that they currently offer. First and foremost, this is an education issue. Thinking “We need a wiki!” is fundamentally a different thought than “We need a tool with which our electronic resources librarians can build an ERMS, because the ‘electronic file cabinet’ approach is not working and we cannot afford Verde or 360,” or “we have many events and no good way to publicize and receive feedback about them.” I suppose this is where Blyberg’s technologists come in (of whom, he intones, there are very few in libraries)–in this scenario, the technologist would hear the need for a home-grown ERMS and think that perhaps a wiki would suffice until money comes flowing in. If librarians made a point of learning about web 2.0 technologies, some of the leaps required to get from need to tool (even experimental tool) could be made without having a technologist on board.

I’m very pleased that our library is doing Learning 2.0. In fact, I’m so jazzed about this, I can’t even tell you. The librarians that work in my organization are ripe for this kind of training, and I expect that once they follow the lessons and learn about the real life applications of blogs, wikis, flickr, tags, podcasts and all the other great stuff we’ll be covering, that they will be better equipped to see not how we can be “EKU Libraries 2.0″ but how we can transform our services to include interactive content that takes advantage of web 2.0 tools.

OK, a reality check

If you read my post from a few days ago titled “Library Agitprop,” you’ll know that I think that OPACs suck, and that I was really angry about that. I had originally written most of that text in my journal while waiting for a flight to wing me home from Electronic Resources & Libraries 2007. I published it on this blog for one reason only–so that I would have to think about it more. I knew that if it were out there and that if people had read it, I would stew about it and be compelled to write a follow-up. There is a lot of change coming to libraries, and rather than railing about it, it’s important (to me) to be a part of it. And, hey, I have to make a living, after all.

The reality is that the ease-of-searching of OPACs has not kept up with A&I database and electronic journal collection interfaces. We should NOT have to pay tens of thousands of dollars for a new interface to plop on top of our OPAC to get it to work better. Our ILS vendors should be writing those and providing those to us. After thinking a bit about it, I think it’s a matter of economics: database and journal package vendors make money from the content that they sell, not from the software through which the content is delivered. ILS vendors, on the other hand, are peddling their software and little else. So, most libraries are stuck with the last version of that software that they were able to afford and are now faced with having to buy Encore, Primo, Aquabrowser or Endeca to make those catalogs more usable–and how many of us can afford to do that?

I am mortified every time I get a research question that involves showing a student how to use the OPAC. The time for having to think about *how* to execute a basic keyword search has long past. Think about the answer to “Do you have the New York Times from 1959?” Well, yes we do, obviously, but now I have to remember how to figure out where it is. It goes something like this:

Step 1: Title search for new york times.


Title browse for “new york times”
Originally uploaded by cindiann.

Title browse for “new york times”

Step 2: I click on #1 in the title browse screen to view all the records that contain “new york times.”


I click “New York Times,” and this is what I get. Uhhhh…..?
Originally uploaded by cindiann.

Step 3: These results appear to be alphabetical. I’ll exercise my shortening patience and look for the Ns.


Here are the Ns, on page 2
Originally uploaded by cindiann.

Here are the Ns, on page 2

Step 4: I don’t see anything that says “1959.” Being a librarian, I know that #19 is the record for a serial because of the open-ended date (1851-). A student might be temped to click on number 20, published in 1960. That’s close.


Clearly, this isn’t right.
Originally uploaded by cindiann.

Clearly, this isn’t right.

Step 5: I return to the title browse screen (probably by clicking “Back,” not the “return to browse” button) and resolutely click on each result that is actually titled “New York Times.” Wishing it wasn’t 3:00 a.m. and that I could ask a librarian. Surely they know.


OK, I finally have the right record. But what does this mean? They only have this month in paper and 1851 in Microfilm…?
Originally uploaded by cindiann.

OK, I finally have the right record. But what does this mean? They only have this month in paper and 1851 in Microfilm…?

If I keep reading the record, I’ll see that Honnold really does have the entire run on microfilm. But who the heck wants to use microfilm? How very James Bond, and not even in a fun way.

Wait, isn’t this the age of the internet? Shouldn’t the library have the entire New York Times via the web? Maybe I should search Google.

So, what do we do? At a recent Library Directors’ Symposium offered by Innovative, a panel of students were very strident about the fact that they never check out books. no one dared asked them if they bothered to search for books, a logical follow-on question. Students turn to Google frequently; should we be pushing Google to index our content? What if the owners of that content (it’s certainly not us) don’t want them to?

Other questions I have:

  • How can we hack at this problem with instruction?
  • Out-of-context instruction sessions to little good. What should we be doing instead?
  • Is the quality of student output suffering if they really don’t check out books?
  • Even though checkout counts are down historically, someone is still checking out books. Who is this, and can we enlist their help?
  • The directional/reference statistics-collecting dichotomy seems outdated; how can we change this so that we get meaningful statistics about what our reference librarians really are doing? Should be doing? (Twitter comes to mind)

Library Agitprop

I first read about Karen Schneider about 10 years ago, when she was a frequent poster to (and perhaps more key, I a more frequent reader of) the web4lib listserv (remember listserves?). I found her frankly irritating and annoyingly loquacious. I cannot remember now why I had this attitude, but it got to where I was hitting the delete key anytime I saw her name.

Funny how things change…

A couple of years ago, I started seeing blog posts, magazine articles and other writings attributed to Ms. Schneider that started to resonate with me. Then she posted the famous (or perhaps forgotten, ymmv) “The User is Not Broken: the Meme Masquerading as a Manifesto.” I don’t think I’ve nodded my way more enthusiastically through a blog post since.

Yesterday, Karen Schneider wrote on ALA’s Techsource blog:

It is both ironic and poignant that librarians are still worrying about “bibliographic control,” after ceding so much of the same to the companies that now rent them journal access per annum at usurious rates, digitize their book collections into DRM obscurity, or sell them ponderous, antiquated “management” systems that on close inspection do little more than serve as storehouses for the metadata specific to the formats of bygone eras, bold days when we saw our central roles as defenders and curators of our cultural heritage.

*blink*

Never have I seen a more eloquent indictment of the ILS as it is today. I was going to wait until I officially leave my current employer before posting something that’s really been eating at me about libraries, but I really feel compelled to weigh in, here.

So-called “next generation” interfaces to today’s OPACs, like Innovative’s Encore and Ex-Libris’ Primo still emphasize print over electronic. Rather they emphasize what is in the catalog over what is in our journal collections (largely available via abstracting and indexing databases, most emphatically not via the catalog). Even with the integration of federated search, the Encore catalog results are displayed first. The results from the A&I databases are drawn in later, and even then it only displays the number of results from each database, rather than details about the results. Someone at Innovative even said that this was on purpose, so that the user would not be distracted by shiny article results (my words). ILS vendors have been doing this for so long that they don’t realize that they’re missing the point nearly completely. Users, particularly student users, but also faculty in increasing numbers in academic libraries don’t want to come to the library to do their research. They don’t care what is IN the library, they want information to be delivered to their desktops. Why else do they start with Google?

As J.R. Jenkins of Serials Solutions said in his recent talk at Electronic Resources & Libraries (and he was quoting Marcia Bates), we are fighting millions of years of evolution here by asking people to do things they don’t want to do, by asking them to do what’s more effort instead of making the things they really need more easily accessible. So, it’s time for my own set of theses on the future of libraries:

1. It is no longer about print. Libraries must figure out a new way to deal with their physical collections that is user-friendly. Deliver the materials to their doorstep, champion portable e-book technologies that are standards-driven and sustainable, something, anything that acknowledges that it’s a pain to schlep to the library.

2. If they want to survive, ILS vendors have to stop listening to librarians and start listening to students. If ILS vendors continue to market and listen to established librarians who largely don’t Get It(TM), they will lose this battle. A library catalog that looks like Amazon.com is not good enough. It might have been a good start eight years ago, even five, but now it’s too little, too late.

3. It’s not about incremental changes and annual releases; it’s about revolution. what we need is the Enterprise-D; what we’ve got is the post-iceberg Titanic, and librarians are convinced that rearranging the deck chairs or buying new, hand-embroidered chair cushions will stop the water from pouring in. Assuming that they even know that there is water; many of them don’t.

4. Federated searching is broken. I hope and pray that someone corrects me on this one. Serials Solutions at least admits its Central Search has shortcomings and tells everyone what it is NOT, but it will still fall far short of users’ expectations, which are fueled not by Amazon but by Google. Federated searching cannot be fixed until applications a) index all relevant content *before* a search is executed, and b) actually search all the content. The idea was that a federated search application sends search criteria to an A&I database, which returns relevant requests *because* it searches the content. Somehow, that didn’t happen.

5. It’s about the users, stupid. It’s as if some librarians are the blind guides in a room full of deaf people. The librarians not only have no desire to “speak” the same language as the users, they are incapable of understanding it. Librarians who aren’t afraid of wikis, who have embraced instant messaging and gaming, who are expert web searchers (and perhaps more importantly, web *consumers*) stand a chance. Librarians who think that the only way to teach students research skills is in the classroom or that Facebook is silly or that fun is a dirty word or that multitasking means having Outlook and Word open at the same time are lost. Simply lost.

6. Google Scholar is also broken. It was a half-hearted attempt not to lose librarians and to give scientists and other researchers some focus, but it just doesn’t cut the mustard. Why? Data. Think about it–Google searches are usually spot-on for two reasons: they have access to all the data on a web page and so take than into account when indexing; also, they were ingenious enough to incorporate linking data into the search algorithms. If a site is linked to by hundreds of thousands of others, it must be good (and if that’s news to you as a librarian, you might as well turn off your computer and go home). I can only assume that those who first did slashdot took this to advisement (not to mention digg).

So, how do we fix Google Scholar? Easy: give them the content. All of it. the indexing, the abstracting, the full text, all of it. Their motto is “do no evil”; what could possibly go wrong? It bothers me very little that publishers and printers gnash their teeth at this idea.

Another factor here that could improve Scholar is cited data. An algorithm analyzing cited data, weighted slightly for currency (varying by discipline), coupled with having all the content indexed and ready before the first search is executed, now that approaches what researchers expect. It’s all on the Internet, right? The cherry on top of this pie-in-the-sky is one-click desktop delivery in the format of the user’s choosing. At no charge to the user, of course.

7. User-generated content is key in the new web context. The only thing cheaper than processing power and PHP scripting these days is disk space. What is a number-one complaint in any library these days? That they don’t have enough staff to do anything new. need to build a database of reference questions and their answers (or where those answers can be found)? Create a wiki and let students populate it. They love to tell stories. let students and faculty review resources.

It’s well past my time to go home for the day, so a rousing and tidy conclusion will have to wait. I will say I need to thank Karen Schneider for this particular kick in the pants. ;)

CiL 2005 – The First Day

Clifford Lynch gave a very entertaining keynote address to open the 20th annual Computers in Libraries conference. He looked back over the “digital decades,” but what I found more interesting was his casting forward and identifying trends that will change and shape the information landscape over the digital decades to come. I’m not sure that his crystal ball is any clearer than anyone else’s, but I like what he had to say, namely:

We are in an era of information abundance rather than scarcity. Remember the days when we paid connect time by the minute and searches by the record? Google’s efforts with libraries will mean that eventually a great deal of public domain works will be available on the net, and publishers are ensuring that more and more complete runs of journals are also available online.

Second, we’re moving from a world of digital surrogates (i.e. bibliographic records) to a world of digital representations that can stand alone (i.e. the full text of an article). Lynch pointed out that this transition has been culturally difficult for many in the library world to make.

Third comes an age of broader authorship. He mentioned The Blog People (who clapped) and wondered if we were heading for an era of popular authorship. Lynch also pointed out that there will be a greater mix of data that was meant to be read by humans with data meant to be read by machines. For example, there aren’t any online maps anymore, per se, but giant geodatabases that can create human-readable maps whenever we need them.

Other sessions I attended today included a session on Electronic Resources Management, during which Andrew Pace from NCSU showed us their incredibly intricate home-grown e-Matrix. Nathan Robertson from Johns Hopkins talked about his work with the Digital Library Federation’s Electronic Resources Management Initiative group, which aimed to create guidelines to speed the development of sorely needed ERM systems. Berit Nelson from Sirsi also spoke about that company’s ERM system; the major difference between Sirsi’s and NCSU’s systems is that Sirsi has endeavored (ha) to shape their existing ILS modules to support ERM (and thereby ensure the longevity of MARC. I think NCSU is more on-target, IMHO.

I also listened to an interesting talk by Northwestern’s Frank Cervone on Institutional Digital Repositories (I finally know what FEDORA is!), and listened to someone from Thomson/Gale whose name I can’t remember talk about collaboration between that company and libraries to create Early English Books and Eighteenth Century Collections Online.

I had mixed success with networking around here today, including down in the exhibit hall for the Cybertour that I gave on OpenURL (which went fine in spite of that). The Hilton’s internet access went down completely for several hours this afternoon, which was frustrating. Obviously all is well again and The Blog People are At It once more.

Another check-in tomorrow! I need to go rescue Ken from the bean, who’s probably a terror by now.

Internet Librarian 2004: "Federated Searching and OpenURL"

Frank Cervone, Northwestern U., presenter

85% of library users surveyed would use a single interface like that at Northwestern (click guest access) if given the option. Notes that not all databases are “federatable”: must use Z39.50 or have an XML gateway/output. FedSearch products are limited to searching 8 databases at a time (product developer at Serials solutions later refuted this—this is a number put out by MuseGlobal based on database connect time and results size). Northwestern’s FedSearch software allows for personalization; one can log in and save searches and favorite databases. Generally, supported searches are title, author, ISSN/ISBN, keyword, and LCSH. Issues with setting up FedSearching include time to configure the databases (very long time); to de-dup (or not); defining searchable collections by subject area (NW has 45 subject areas).

What we could use: explore Fed Search for general databases to improve upon the interfaces that we don’t like; use OpenURL to link content from there.

Challenges: Fed Search is a dirty word; lack of expertise; shortage of time.

Internet Librarian 2004: "The Library Web Site meets About.com"

David King Kansas City Public Library told us how KCPL uses a unique approach to create subject guides for their web site that are modeled after the guided approach of about.com. They analyzed the links that they were maintaining and found that the only two kinds of things being used were the local information links and their subscription databases. Content about a single subject was distributed across many web pages: the catalog, web site, readers’ advisory pages, calendar pages, book lists. They pulled them all together, added longer articles written by their house experts, and added headlines fed from other sites (using RSS) to round out the content. Example: Harry Potter info page, the computer guide, or the “how do I?” guide.

Technical stuff: use an SQL database and Cold Fusion. The articles and events are “modules,” each of which is assigned a category. They use a homegrown editing system that looks a lot like our subject research guides. A guide can be “live” or not—only live guides are displayed on their web pages. They used a homegrown statistics gathering program to analyze the site before it was redesigned.

What we could use from this idea: Solicit content from subject specialists, but ownership and responsibility for the web site itself rests with a content manager.

Challenges: inertia of status quo; perfectionism; shortage of time to implement

IL 04: faceted metadata

“Facets” are attributes of an object that are 1) mutually exclusive and 2) sufficient to describe that object. Facets allow for the creation of a controlled vocabulary.

Metadata is “data about data,” or everything but the text. Sources of metadata can be explicit (tags or properties); intrinsic (the author or date may be noted in the text itself); derivable (file size, file format); and external (category, topic, subject headings). There is metadata in MARC records (author, title, subject, main entry, publisher, place, etc); Dublin Core metadata allows for the description of (digital) objects other than books (source, audience, coverage, format).

Faceted metadata is various aspects of an information object that are not just topical. We must expose facets to enable browsing. “Facets” of a shoe could be: size, shape, color, style, model, brand.

Library catalogs based on MARC records often search like other web sites—with a command-line interface or a web form that presumes that the searcher either knows the command syntax or can understand what the fields are (not to put title in the author field, for example).

A full text search, on the other hand, is too broad, returning hits without context—it’s difficult to know without examining individual records if a search for “U2″ retrieves items related to the Irish rock band or the spyplane from which the name came. Displaying metadata of the items retrieved in terms of its facets gives context and allows the user to switch between searching and browsing. For example, an artist search at towerrecords.com will show not only the resulting matching albums but a set of categories that allow browsing of related items by genre, price, year, format, and other “facets.” Imagine a library catalog that displays not only the matching titles for an author search of “Joyce, James” (typed in as james joyce, even) but a list of categories along one side that allows the user to browse the results and related titles by type (primary resource, criticism, essays), publication date, location, availability, language, length, or even the public librarian’s dreaded question: size and color.

Read more about faceted metadata and link to related resources at Avi’s site, Searchtools.com.