RE: HM.Frameworks: Physical Description

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

KindaLike decided by what over whom?

 

Will you bet someone's life on that because in a public safety system,

you may be doing that? Is KindaLike that ranks identity of a person

likely to perform better than a candidate list of best matches

based on string comparisons of data?  Eventually, it results in a list of

matches from which a human with the authority to choose

chooses.   Keep in mind that when selecting humans out

of a lineup, the ability of the lawyer to contest the selection is

based not on the absolute refutation of a choice, but one

the possibility of error.   What you can use kindaLike for

is to create a line up based on the similarity in demographic

data and the mugshots.

 

AI is neat stuff and in expert systems, sometimes by dint

of arduous and expensive effort, useful.   Historically, it

didn't work better than having humans do it themselves.

It was useful for capturing expertise that was reasonably

well-described and mostly static.  I am unconvinced that

the semantic web changes any of this and therefore,

no secret, not as interested in it as the data descriptions

given that the vast (say kindaLike 99.44%) of systems

use data transactions.

 

So, I question the application.  Physical Description

seems to be an unlikely candidate depending on the

application and I am not comfortable with creating

requirements for which I don't have a matching system.

Len Bullard
Intergraph Public Safety
clbullar@ingr.com
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h

-----Original Message-----
From: Kurt Cagle [mailto:cagle@olywa.net]
Sent: Tuesday, August 28, 2001 12:50 PM
To: humanmarkup-comment@lists.oasis-open.org
Subject: Re: HM.Frameworks: Physical Description

I'm noticing on the discussion on arcs (kindaLike, etc.) that we seem to be moving into a Fuzzy Logic (FL) realm. It occurs to me that it may be worth either labeling or supporting arcs with an intensity indicator as well [0,1] that would provide a mechanism for correlation across a specific operator. For instance, if you have an arc operator of "Like" then a like intensity of 0 would correspond to no correlation (the arc between the two entities implies no correlation) while an intensity of 1 would imply total correlation (1 is a synonym for the other) and an intensity of 0.7 would indicate that there is a certain amount of correlation (the two are similar but not the same). While for some of the arcs being dealt with these intensities are largely subjective, the same can be said for HumanML in general.

 

I'd push this approach for a couple of reasons. First of all, it makes it possible to follow correlative links into a hyper-linked environment, making it feasible to provide a point at which point information ceases being reliably relevant.  For instance, if you had an arc:

 

     Like at 0.7

A ---------------------> B

 

and another arc

 

     Like at 0.8

B ---------------------> C

 

then you can view the arc A->C as being similar at 56% (which could be interpreted as there is about a 1 in 2 chance that the two nodes are highly similarly, or more accurately that there is a 56% correlation between A and C). You could then follow a grove of nodes pruning those that fall below a certain threshhold (say 0.1 correlation) as no longer being relevant to the current arc. This also makes it possible to perform other analytical metrics (Bayes, for instance) or to determine the mean relevance of a tree to a given node.

 

This is just an observation, do with it as you will.

 

-- Kurt