Representations of documents with Icons or glyphs

Genny Engel gen at dla.ucop.edu
Sun Aug 10 02:56:42 EDT 1997


Look into the information browsing literature in human-computer
interaction.  There are some abstracts of recent BayCHI talks on related
topics at http://www.baychi.org and I'm sure there's much more in CHI
proceedings and the SIGCHI Bulletin.  Try looking for Focus+Context for
some relevant work.  HFES may also have some related stuff.

Genny Engel
MELVYL System User Services
and BayCHI News editor
gen at dla.ucop.edu

On Thu, 31 Jul 1997, Alain Vaillancourt wrote:

> 
> Hello!
> 
> In the field of Scientific data visualisation multivariate numerical 
> data is sometimes expressed with an icon or a glyph.
> 
> Sometimes this glyph is mapped onto a 2D or 3D representation of the 
> experiment, as in the work of Wittenbrink, Pang and Lohda.
> 
> Sometimes the icons or glyphs are presented without any background, 
> in large arrays permitting users to spot similarities between them, 
> as is usually the case with Chernoff faces or their descendants, Flury-
> Riedwyl faces.
> 
> I am interested in doing research on ways non-numerical data (such as 
> an article's author, title or subject) can be expressed with icons or 
> glyphs, in circumstances where they are presented with no background, 
> but also in circumstances where they are mapped onto significant 
> infromation visualisation images like Kohonen maps or other self-
> extracting maps.
> 
> I went around the information Visualisation projects collected on the 
> Cyberstacks sites and I found none that looked in any way at the 
> question of iconic or glyphic document representation.  The most 
> sophisticated ones used colored spheres to represent documents.  Which 
> is a long way off from data expressed in Chernhoff faces or typical 
> data glyphs.
> 
> Is anyone aware of any research being done on the topic?
> 
> I have the impression I am about to do some rather heavy 
> trailblazing, since mapping textual or conceptual data on glyphs 
> seems to be radically different from mapping numerical data.
> 
> It is very interesting though.
> 
> Thank you for your directions, if any
> 
> Alain Vaillancourt
> 
> cxv3 at musica.mcgill.ca
> 
> or
> 
> ndgmtlcd at gslis.lan.mcgill.ca
> 
> 4320 Kensington
> Montreal QC
> H4B 2W3
> 



More information about the Web4lib mailing list