Newsgroups: comp.lang.prolog
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!pipex!sunic!erinews.ericsson.se!ericom!acrosby6.ericsson.se!etmerei
From: etmerei@acrosby6.ericsson.se (Erik Reitsma)
Subject: Re: New revolutionary Prolog derivative
Message-ID: <1994Nov30.123522.195@ericsson.se>
Sender: news@ericsson.se
Nntp-Posting-Host: acrosby6.ericsson.se
Organization: Ericsson Telecommunication B.V., Rijen, The Netherlands (ETM)
References: <3bgkme$q6e$1@mhadf.production.compuserve.com>
Date: Wed, 30 Nov 1994 12:35:22 GMT
Lines: 37

In article <3bgkme$q6e$1@mhadf.production.compuserve.com>,
Colbert Philippe  <74671.1062@CompuServe.COM> wrote:
>Years ago I have read a book called IMPLEMENTING PERSISTENT PROLOG. 
>This book shows how to write a Prolog compiler.  This exercise had me
>thinking for about 4 years.  Now I think I can greatly improve on the
>limitations of Prolog.  I would appreciate if some knowledgable
>academic would be willing to corresponding with me to discuss my
>idea.  Please mail me back.  My idea goes as follows, the popular
>implementations demonstates the power of deductive reasonning based
>of 1st order logic (logic programming).  But in order for prolog to
>be more useful it needs to work with things other than string
>representations.  My innovation would be to have Prolog work with
>concepts rather than just strings.  The concepts would be represented
>by graphs.  I have designed a "Very Special" algorithm that can
>decompose a graph (decomposing a concept) by abtracting...(can't
>reveal everything..).  I will implement this algorithm in a prolog
>compiler.  This way prolog can take a vague concept, decompose it
>and match to the knowledge base (binding).  This is very powerful
>stuf !  Please respond anybody !!!
>
By concepts and graphs, do you mean conceptual graphs as defined in
Sowa's "Conceptual Structures", esp. chapters 3 and 4? If you
are planning to implement the algorithms described there in Prolog,
I certainly would be interested in your results.

Your description is a bit vague, but it sounds a bit like Sowa's
book. Perhaps you should read it. (ISBN 0-201-14462-7, Addison-Wesley,
not sure wether it's still available though, might be out of print...)

Good luck!

Erik Reitsma
--
     - - - -   
    /    ____  
    __  |    |  " = Ugchelen, Pearl of Gelderland (NL) 1986
   / /  _\   |  
