[
Lists Home |
Date Index |
Thread Index
]
> That's presumably how humans cope with the problem of distinguishing
> "tables" from "tables". People know better than to do this, but
> most could understand a sentence like: "We need to table the discussion of
> how to format the table for our catalog of tables." No URI's needed!
So the question may be more-- when do we need identity? When we need to
understand the sentence we need to be able to identify each of the "table"
words in the above sentence, but not to pass it around in an email-- it is
just text. In my brain at least the above is parsed as follows:
1) upon encountering the first table mentally call up the definition of
"table" I come up with several
2) determine usage (the first is a verb)
3) determine defininition based on context
In cases where I don't know the word-- the second two may suffice. I wonder
if this can be used in software. When we begin parsing we have a list of
recognized vocabularies, (which may each include a "definition" for a table
element)-- once we have identified the possibilities then determine usage
(root, container, empty) and finally if there is still overlap determine by
context. Of course this is application level code. In a FurnitureML renderer
suppose it encounters a
<MartianRover> in the position that a <table> would normally go. It may not
understand the term MatianRover but it could assume based on usage and
context that it is *some* kind of furniture. It could simply draw a box in
place where it would normally draw a table.
But is this an improvement?
Jeff Rafter
Defined Systems
http://www.defined.net
XML Development and Developer Web Hosting
|
