Dear DC Colleague,
I copy below an announcement of a study which I'm piloting within my own
research community in Human-Computer Interaction (HCI), but which as a
metadata strategy hopefully has broader application. There's a brief intro,
and then links to a short article and the pilot metadata form.
It takes as its point of departure something like the Relation field in DC,
and proposes, in the DC spirit, that different scholarly/scientific
communities now need to specialise this into sets of scientific/scholarly
relationships of particular importance to them.
I focus on formalising relationships rather than content descriptors since
it is more likely that an evolving, relatively young research/design
discipline can agree on a WAY to conduct discourse before it is likely to
be able to agree on some kind of master classification scheme for
describing document content. Loading the links (relaitons) rather than the
nodes (resources) with the metadata in this 'semantic hypertext' also
allows the same document/idea to be recast in multiple ways by different
researchers, a key requirement.
If any of you have comments on the approach I'm taking, I'd be delighted to
hear from you. I'd also be interested in pointers to other communities who
are extending DC in similar ways.
Simon
-------------------------------------------------------------
It's been said many times in recent years, and it's still true:
The Net, particularly the Web, provides an unprecedented opportunity in
scientific history to locate, interconnect and analyse ideas and documents.
ButŠ
The Web is becoming a more chaotic place by the day. As the signal to noise
ratio gets worse, research communities need better support for tracking
developments and finding relevant documents.
It is currently impossible for search engines (Web or otherwise) to answer
complex questions commonly posed by researchers such as the following:
* are there distinct schools of thought in this field?
* what impact did this evidence have?
* who is currently tackling this applied problem?
* has anyone built a system based on this theory?
* has anyone applied this theory to other fields?
Is it possible to conceive of a 'linga franca' for summarising key
scholarly and scientific relationships between research documents which
could provide a representational infrastructure capable of supporting such
queries? What if we could agree that - within HCI research - there is a
fairly stable set of inter-document relationships of interest and
importance that would be worth encoding in a form detectable, and
visualizable, by software agents?
If interested, or at least mildly intruiged, read on!...
AN INVITATION
I'd like to invite you to participate in a pilot study to evaluate a
"metadata scheme" for describing HCI research publications. This provides a
set of scientific relationships which can be used to conceptually link a
document to others in the research literature.
The "HCI Knowledge Web" pilot website is at:
http://kmi.open.ac.uk/sbs/hciweb/pilot.html
from where you can download a short article (summary below) setting out the
rationale for this project, and a template form with the proposed metadata
scheme and examples.
I invite you to try describing 1-2 of your own HCI publications, and
sending me the form and your feedback.
Thanks,
Simon
...........................................................................
http://kmi.open.ac.uk/sbs/hciweb/Interfaces98.pdf
Evolving the Web for Scientific Knowledge:
First Steps Towards an "HCI Knowledge Web"
Summary: In this article, I consider the challenge of building a Web-based
infrastructure for scholarly research which moves beyond the basic
dissemination and linking of documents, to support more powerful searching
and analysis of the cumulative knowledge in the literature's documents.
Taking the HCI research community as an example, the goal would be to
enable HCI researchers to search for interesting documents and phenomena,
and discover previously unknown but conceptually related research, for
instance, other groups addressing persistent problems in the field, the
structure of debates, or when and how new theoretical perspectives began to
make an impact. I propose that focusing on the scientific relationships
between documents is important, and has advantages as the basis for a Web
metadata scheme to enrich the HCI community's Web.
To appear in Interfaces Magazine, British HCI Group, December, 1998.
__________________________________________________________________
Dr Simon Buckingham Shum Knowledge Media Institute
The Open University Milton Keynes, MK7 6AA, UK
Alias Email: [log in to unmask] http://kmi.open.ac.uk/sbs/
Tel & Vmail: +44 1908 655723 Fax: +44 1908 653169
J.Interactive Media in Education http://www-jime.open.ac.uk
D3E Interactive Web Documents http://d3e.open.ac.uk
British Human-Computer Interaction Grp http://www.bcs.org.uk/hci
__________________________________________________________________
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|