Re: storing info in XSL-FO: new issue? [was: Draft TAG Finding:...]

At 10:28 AM +0200 8/20/02, Robin Berjon wrote:

>The part I fail to understand is "even without agreement". How is 
>that going to make more sense than a vocabulary that has agreed upon 
>semantics such as MathML? Are we talking about a musician or about a 
>sewing machine model?

Now how do we process and make use of the meaning that is in the 
message, given that we don't have a prior agreement? To answer your 
question, how do we figure out whether we're talking about a musician 
or about a sewing machine model? In this case, it might require some 
human intervention to decide that we were talking about the singer 
instead of the company. In the future, maybe not. I've seen at least 
one knowledge based system that could probably figure out which one 
we meant.

But even without any sort of AI, a human can look at the data, 
reverse engineer the XML, and write software to process the 
undocumented format, without any prior agreement. The process 
generating the data need not know that this is happening. Our results 
will be imperfect, and will occasionally fail, just as they do in the 
real world with data that is transmitted under an agreement, but 
properly handled those results can still be useful. We may need to 
occasionally flag problems for human intervention, but a lot of 
details can be automated.

I do this sort of thing frequently, with both XML and non-XML 
formats. In practice it is *much* easier with XML based formats 
because they contain more information. It is more useful to know that 
Madonna is a singer than a list item. XML markup contains more 
information than HTML markup in most cases.

| Elliotte Rusty Harold | | Writer/Programmer |
|          XML in a  Nutshell, 2nd Edition (O'Reilly, 2002)          |
|                  |
|  |
|  Read Cafe au Lait for Java News:      |
|  Read Cafe con Leche for XML News:    |

Received on Tuesday, 20 August 2002 09:30:29 UTC