Adaptive hypertext and hypermedia for information retrieval

[Prentiss Riddle]

Steve Ryan (s.ryan at library.usyd.edu.au) is looking for people who are
using adaptive hypertext and writes:

> Date: Sun, 30 Jun 1996 17:04:57 -0700
> From: s.ryan at library.usyd.edu.au (Steve Ryan)
> To: Multiple recipients of list <web4lib at library.berkeley.edu>
> Subject: Re: Adaptive hypertext and hypermedia for information retrieval
> 
> In answer to Prentiss' question I'll quote from a paper by Peter
> Brusilovsky (http://www.edfac.usyd.edu.au/projects/ah/Brusilovsky.html)
> 
> " ... Users with different goals and knowledge may be interested in
> different pieces of information presented on a regular page and may
> use different links for navigation. ...  A way to overcome this
> problem is to use the information about a particular user,
> represented in the user model, to adapt the information and links
> being presented to the given user. I call it adaptive presentation.
> Knowing user goals and knowledge, an adaptive hypermedia system can
> support users in their navigation by limiting browsing space,
> providing adaptive comments to visible links or just suggesting most
> relevant links to follow. I call it adaptive navigation. "

Thanks.  The definition is helpful.  I note that another buzzword under
which this sort of thing often falls is the "user profile".

One problem I have with the (admittedly limited number of) user
profile-based systems I've seen or heard about is that they often seem
to be rather naive in their assumptions about the monolithic nature of
the "user".  The simplistic user profile model fails to take into
account that the user has varied and conflicting interests (e.g., jazz
music and Java programming, but not necessarily the intersection of
jazz music and Java programming); that the user may have different
roles (i.e., Java programmer on the job and jazz lover after hours);
that the user's interests and expertise change from moment to moment
(i.e., an introduction to Java on Monday and an expert-level discussion
of Java security issues on Friday); and that the user may act as a
proxy for others whose needs do not match the user's own profile (i.e.,
researching COBOL programming or heavy metal for a friend, and the
problem of how to compose a profile for a reference librarian).

In my opinion, in order for profiles to work, the user needs to be in
complete control of the profile, including the ability to set the
profile aside and take up a new one.  Perhaps a good profile system
would use a metaphor like "hats" and allow the user to maintain a
closet full of "hats", choosing the one that fits best at the moment.

But putting the user back in control creates issues of complexity in
the user interface and may give up the alleged efficiency gains (by
bounding searches, etc.) that claims to be an advantage of adaptive
systems and user profiles.  In other words, an adaptive system with the
user in charge quickly begins to resemble the general-purpose tool
which the adaptive system was supposed to replace.


That said, the adaptive hypermedia systems I've run across on the web
fall into the "tell me what you like and I'll tell you more stuff like it"
category of tools.  In my experience, none of them work very well.
   
   Firefly (aka HOMR aka Ringo): matches movies and music interests
   http://www.agentsinc.com/

   Webdoggie (aka Webhound): matches interesting web sites
   http://webhound.www.media.mit.edu/projects/webhound/www-face/
   
   The Similarities Engine: matches musical interests, working on web sites
   http://www.ari.net/se/

I'd be interested in hearing about more serious applications.

-- Prentiss Riddle ("aprendiz de todo, maestro de nada") riddle at rice.edu
-- RiceInfo Administrator, Rice University / http://is.rice.edu/~riddle
-- Home office: 2002-A Guadalupe St. #285, Austin, TX 78705 / 512-323-0708