On Sun, 30 Dec 2007, John Black wrote:
> > The other big attraction of CGs to me is the attractive possibility
>> that the graphical display of CGs may facilitate the solution of
>> certain types of reasoning problems. I am particularly interested (for
>> no particular practical reason, just am) in the possibility of
>> reasoning about the kind of knowledge problems described in "Reasoning
>> About Knowledge" by Fagin, et al.
> > http://mitpress.mit.edu/catalog/item/default.asp?ttype=2&tid=8240 (01)
Fair enough, but that is not particularly graphical in nature. (02)
> >
>>
>>[Person:*j]<-(KnownBy)<-[Knows:[[Person:*h]<-(KnownBy)<-[Knows:[[Person:*j]<-(KnownBy)<-[Knows:[Cat:Yojo]->(Named)->["Yojo"]]]]]]
>>
>> If I did this right, it means: a person *j knows that a person *h knows that
>> he, *j, knows that Yojo the cat is named "Yojo". Textually, it gets very
>> complicated to look at. Would it be simpler for a human if this was visible
> > graphically?
Chris Menzel <cmenzel@xxxxxxxx>: (03)
>That's of course a matter for psychological research. There's been a
>fair amount of work on graphical reasoning where an answer to the
>question might be found. (04)
And the answer is, yes, for humans, if the display is done right and
the logic isn't too complicated. Put more delicately, the way that a
logical problem is presented can make a lot of difference to how easy
it is for humans to solve it. But this depends on things like how
familiar the human is with other notations and ways of displaying
problems, and it is very case-sensitive. Some people are much less
visually oriented than others. (05)
> > And does it make any difference to a machine?
>
>None, surely, with respect to all the well-known theoretical limitations
>on automated reasoning, as those limitations are all
>representation-independent. (06)
Right. One has to remember that the reason we humans are so good at
visualization is probably that a very large fraction of our cerebral
cortex (around a quarter of it) is entirely devoted to processing
information from our retinas. It is this, rather than anything
intrinsic to visual sensory modalities themselves, which accounts for
the utility of image-like representations to us. But computers aren't
built that way, don't have retinas, and if you examine how they
represent "images", they use much the same representational
techniques that they also use to represent "textual" information. The
most efficient and powerful machine reasoners depend on such highly
inhuman abilities as extremely rapid structural matching mediated by
sophisticated hash-coding schemes, which depend only on the abstract
graph structure of the information, not on how it is displayed on a
screen. (07)
Pat (08)
--
---------------------------------------------------------------------
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 (09)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (010)
|