W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > June 2007

Re: RE: RE: [hcls] A map of the Semantic Web for life science and health care

From: Pat Hayes <phayes@ihmc.us>
Date: Fri, 22 Jun 2007 23:47:57 -0500
Message-Id: <p06230903c2a250a53054@[]>
To: samwald@gmx.at
Cc: eneumann@teranode.com, public-semweb-lifesci@w3.org, VKASHYAP1@PARTNERS.ORG

>Hello Pat,
>>  None of this seems to be best conveyed by a
>>  vaguely metaphorical
>Actually, it a very clearly defined metaphor 
>that also relates to metaphors in existing 
>publications ("bioinformatics nation"). If you 
>have a general dislike of metaphors, I might 
>understand that, but I don't think the specific 
>metaphor I use is vague.

I was imprecise. My intended point is that the 
spatial-map metaphor can be, and has been, used 
to convey a large number of very different 
metaphors. The only way to understand which one 
you intend by it in this case is to read some 
text which explains it. By that time, however, 
the pedagogic utility of the spatial metaphor has 
largely been lost. And I don't think this use of 
it is very convincing or useful. The most salient 
aspects of a real map are direction and boundary 
shape, neither of which have any meaning at all 
in your diagram.

>  > diagram drawn using
>>  arbitrary rules,
>The rules are not arbitrary at all.

They are precisely arbitrary, in the sense that 
they could have been otherwise. "Arbitrary" is 
not a derogatory term.

>Please read the text in my mail or on the wiki 
>page. Looking only at the picture (which is a 
>mere draft of the graphic style) does not tell 
>you much.

Which is exactly my point. If it really were a 
map, looking at it would be sufficient to 
understand it immediately.

>  Most of the features convey meaning, those that 
>do not have an ergonomic function (e.g. slightly 
>different shades of colors make it easier to 
>distinguish areas).

Yes, but there is a principle of GUI design which 
you are violating here. Some of the visible 
structure is meaningful (although the meaning is 
sometimes quite opaque) but other aspects, 
equally visually salient - in fact in some cases 
more salient (color, direction and shape vs. 
connectivity) -  are not. But there is no 
*visual* clue as to which aspects are meaningful. 
In this case, the meaning is hidden rather than 
made manifest. It is actually *harder* to 
understand a graphic like this than it is to 
understand a purely symbolic description of the 
same information.

>  > Why not just SAY all
>>  the above, for example by listing the ontologies
>>  and saying what they have in common?
>The map should show the connectivity of the 
>resources. While it is true that the current 
>connectivity could be communicated with a simple 
>table (since we are not that far with aligning 
>and linking our ontologies/data), I am 
>optimistic that this will change in the future. 
>Since the 'map' should ideally be updated with 
>future developments, we will (hopefully) reach a 
>degree of connectivity that cannot be 
>ergonomically represented in a table.

We have found that very complex connectivity 
graphs are often more ergonomically represented 
as a table. It depends on what exactly one wants 
to do with them, what information is most 
important, and how tangled they are when drawn. 
For example, a highly nonplanar graph with many 
line crossings is a very poor way to present 
connection data when the task is to determine all 
the nodes that are linked to a given node. Also, 
a table-based GUI can often be the most 
ergonomically effective way to input connection 
information as typed node-arc-node triples, 
partly because it can be done entirely using a 

>  > About all I
>>  can get out of it is that there are some
>>  ontologies which are in some sense about some
>>  topic areas. That could be said in a few lines of
>>  English, or a small table.
>It is obvious that you did not read the text on the wiki page, sorry.

But surely, if I have to read the text in order 
to understand the graph, then does that not 
rather weaken the case that the graph is a good 
way to display the information?

>  > But I won't rain on this particular parade again.
>>  Y'all go ahead and draw pictures.
>It is a shortcoming of many great scientist that 
>they lack the motivation to invest time into the 
>communication of their science to people outside 
>their knowledge domain, or even to the general 

No doubt. I am not a great scientist, so would 
not know. However, I do work closely with people 
who study the ergonomics of various forms of 
graphical and diagrammatic display, intended for 
use by people varying from kindergarten-age 
children to engineers; and I study graphical GUIs 
of various kinds. I am not terribly impressed by 
the often assumed advantages of graphical 
metaphors, which need to be used with skill and 
care in order to be really effective.   BTW, I do 
not claim to possess the relevant skills, but I 
know that there are people who do.

OK, REALLY no more from me on this topic.


>Doing that requires making compromises, 
>metaphors, slightly ironic illustrations and 
>other things that are not part of 'hard' 
>science. This is rather unfortunate, since it 
>hinders widespread acceptance, recognition and 
>understanding of most scientific developments 
>outside a very limited field of experts.
>Matthias Samwald
>Yale Center for Medical Informatics, New Haven /
>Section on Medical Expert and Knowledge-Based Systems, Vienna /
>Psssst! Schon vom neuen GMX MultiMessenger gehört?
>Der kanns mit allen: http://www.gmx.net/de/go/multimessenger

IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
Received on Saturday, 23 June 2007 04:48:11 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:20:28 UTC