Tuesday, December 01, 2009

More thoughts about ebooks and online reference collections

My last two posts have been a mash of various conflicting thoughts about organizing and presenting the Materials Formally Known As Reference in an online environment.

What I neglected to say in those posts is make note of what most academic libraries are doing about the situation at the moment. Most academic libraries still hand-select the material that they present as 'reference' and tend to they divide these pages by format. There use boxes or tabs or lists and more frequently, a combination of all of the above).

I know there was more I wanted to write about this but I put off writing and these days if I don't capture my thoughts quickly, my thoughts get indignant and storm off, leaving without a tip.

So please bear with another half-formed idea.

One of the difficult things about the digital library is that each library is really just a collection of various access rights over different databases. The academic library website is really just a bunch of links going to a whole bunch of different search boxes. In fact, almost every collection -- including the catalogue - must be entered through such a search box. It is as if its a magic keyhole. And it makes it difficult to make a sense of a 'whole'.

But what if the library website was more like Wikipedia. Wikipedia gives you the feeling that its a "space." Wikis are like that. A good one is populated. A bad one is empty. And even though Wikipedia is as database driven as the library catalogue, the human-readable names of pages and the ability to make meaningful connections to related pages, gives it more of a feeling of a book than of a 'database'.

What if we had a library website that created and named a web page for every Library of Congress Subject heading (and free floating subdivision). Within this page, there were would be a dynamically generated list of library owned or licensed books that fell under that subject as well as a section that brought back the abstracts and links of articles by a keyword match. Librarians, and hell, maybe even users, could supplement each subject page with relevant materials from the library catalogue or the web.

Currently, most academic library websites have a page for each department on campus and some libraries have expanded this list to each university program. There have been steps created to create a library presence for every course offered on campus. The amount of effort to hand-pick resources for every subject offered in a university is daunting, so why not start considering dynamically selected materials?

1 comment:

pzed said...

Drupal has this concept of nodes: "the generic term for a piece of content on your web site." (Since I'm commenting, I won't link, but all quotes are from the Drupal site, somewhere.) The node module allows "Modules and scripts [to] programmatically submit nodes using the usual form API pattern." Which means nodes on a website could conceivably be added or edited by triggering a cron job that either requests info through an API or reads a feed.

We think of nodes as being equivalent to posts/pages, but it seems they can be much more: "By using modules such as the contributed Content Construction Kit (CCK) module, the core Taxonomy module, and the contributed Location module, you can add fields and other properties to your nodes."

Can we imagine a node for each reference title (and each item on the a-z list as well, which is reference too of course)? Could these nodes come from a feed generated by one or more sources? Most of our ebooks are supposed to be in Conifer (aren't they?) but Conifer only has one RSS generator: bookbags!

But Conifer does have a public API, although it's not well documented. There may be a way to pull a specific set of ebook records and call it Reference.