3D (VRML) Interface to IRCAM's Multimedia Library

[JQ Johnson]

Jerry Kuntz asks the right question about simulations (VRML or otherwise)
of a library physical space:

>what advantages does VRML add to a library navigation interface?

Let me suggest a few potential ones off the top of my head.

First, and easiest, is a comparison with a "simply rendered floor plan."
There's a lot of good cognitive psych research that shows cognitive style
differences between people who use plan and route knowledge to do
wayfinding.  A plan is a rather poor tool for the people who need a route,
and vv.  So it's at least plausible to believe that providing virtual
routes to resources in a complex 3 d space (most libraries) would be a
useful navigational tool.  Here the task is finding specific resources
that have already been identified by some other means (such as a
hierarchical textual catalog or a web page that says "we have a scanner in
our Information Technology Center" but doesn't say how to find the ITC).
Unlike a physical walkthrough, a virtual one also allows easy perspective
shifts (so I could switch between a route view and a birds eye view that
is quite map-like, or zoom out to find out where the NA books are filed
then zoom in on a particular display case in the collection of library
scale models).

A realistic virtual world also provides a good browsing interface where
one's browsing depends on some physical characteristic of the space.  I
frequently browse a shelf and find interesting works near (defined by LC
number, since that's how we shelve books, or by alpha title, if that's how
current periodicals are shelved) a work I started looking for.  Or,
perhaps a more common problem, consider finding a good study location
somewhere in the stacks.

Third, it's plausible to believe that, by humanizing the presentation,
providing a realistic virtual world will increase the comfort of people
who are not regular library users, even if it doesn't provide specific
information.  This is the domain of the "virtual library tour".  It's also
one reason why we include photos of our facilities in our marketing
literature.

Fourth, a VRML world is particularly good if the items in the "collection"
have salient physical properties that can be represented in the world, as
is typical in museums and many special collections.  A book is usually
just a book (consider how little we care about those MARC 300 fields!),
but suppose the collection is one of artifacts or service points -- the
VRML world could contain a thumbnail view of the item in ways that are
very hard to incorporate in a traditional textual OPAC.

Fifth, a VRML world is a very simple representation in some ways -- it's
not abstracted far from the physical reality.  That can make it easier to
add new analogous features later.  Where does one place a link to the
email reference service in an OPAC?  One doesn't; it doesn't fit.  But
there's a natural place to place a connection to the virtual reference
service when one is modelling a library with a VRML world -- the reference
desk.

I don't personally find that a realistic VRML model of a library is useful
to me as a patron, at least not until it is extended with abstracted
information that is NOT available in an actual walkthrough.  And
(technical aside) I'm not wild about VRML as a  modelling language.  But I
could easily see the utility of a VRML-like model that incorporated
additional information not obvious in a physical walkthrough: highlighting
target items/destinations using color, offering multiple simultaneous
perspectives (e.g. showing map PLUS current position and route PLUS
textual description), offering multiple views (a mobility-impaired patron
might want to see a view that emphasizes the sizes of corridors and
whether books are on shelves that are within reach from a wheel chair; a
laptop user may want to see a view that emphasizes where the carrels with
10baseT wiring are to be found; depending on task one might want views of
individual shelves where one can toggle between a complete shelf list and
one showing only STATUS=AVAILABLE, etc).

Bottom line is that one should consider such presentations as experiments
with potential benefits, and not dismiss them as frivolous.  But we should
also not run out and implement VRML models of our own libraries tomorrow!

By the way, while we're thinking about star trek technologies, note that
the cost of computers in fabric or plastic is falling fast enough that we
might within a decade or so be able to afford replacing barcodes and
magstripes in books with fairly powerful computers.  Imagine if that
computer-on-a-cover included wireless networking that allowed our VRML
world (not to mention our circ system) to track the location of every
single book in the library...

JQ Johnson                      Office: 115F Knight Library
Academic Education Coordinator  mailto:jqj at darkwing.uoregon.edu
1299 University of Oregon       phone: 1-541-346-1746; -3485 fax
Eugene, OR  97403-1299          http://darkwing.uoregon.edu/~jqj/