ANSI standards and knowledge representation

Erik Sandewall <>
From: Erik Sandewall <>
Date: Thu, 18 Aug 94 17:41:10 +0200
Message-id: <9408181541.AA00526@pan23>
In-reply-to: Timothy Finin's message of Thu, 18 Aug 1994 09:56:21 -0400 <>
Subject: ANSI standards and knowledge representation
Precedence: bulk

Tim Finin writes

>  > Additionally, I propose that the typographical basis should be LaTeX
>  > and not ASCII. In other words, rather than defining expressions in the
>  > language as sequences of ASCII characters that can be read by
>  > COMMONLISP, they should be LaTeX expressions which can both be
>  > presented in normal "textbook" style, and can be read by specialized
>  > parsers.  That overhead is certainly acceptable.
> Interesting idea.  Wouldn't SGML be better for this though?  SGML was
> designed to capture the structure of a document (which is what's
> important here) rather than the lower level presentation information.
> There are good tools for converting SGML to other document languages,
> such as HTML, LaTeX, RTF, etc so presentation per se is not a problem.
> There is a nice parallel here -- SGML is a kind of interlingua for
> document descriptions which we would then be using as a basis for
> encoding an interlingua for knowledge.

My major reason for proposing LaTeX was as a way of obtaining typography
for real formulae. Document structuring may also be important, but 
I would imagine using explicit algol-style codewords (begin module, etc)
even in the final formatted text.

Although there are some arguments in favor of SGML (formal standardization,
usage in WWW, etc), it suffers from
**  Horrible syntax for formulae (at least when I last looked at it)
**  In general, it seems to be have been designed as an interchange
    language between WYSIWYG formatters, whereas LaTeX was designed
    to be input and edited by people. I like to have the hands-on contact
    that the LaTeX source text offers.