dave@umcp-cs.UUCP (Dave Stoffel) (03/29/85)
I am interested in any information, comments, or questions
on the subject of using Prolog in text processing applications.
These applications might be document management, information
retrieval, or story understanding. Word processing is probably
not an application of interest, though I am always open to new
ideas.
I work for a company which has developed a full-text search,
storage, and retrieval software package, not unlike text search
packages available on mainframes, for the IBM PC XT/AT. This
product does not build an inverted file, though it does build an
index. It is very fast, both in searching and building, though
particularly in building. This software package also runs on Unix
machines, though it is not yet available as a product on any.
I don't quite know yet what I want to do, but I do know I
want to explore the oppurtunities suggested in paragraph
1 by building an interface between a prolog interpreter and our
text search product. I would like to solicit any info, suggestions,
criticisms, questions or whatever on the nature of such an interface.
We have developed the search product quite modularly; there is a
search engine, an embedded boolean query language, and a window/menu
subsystem. I have already interfaced a prolog interpreter to the
embedded query language by defining an evaluable predicate `query(S)',
where `S' is an atomic search string, but this hasn't proved too
interesting yet.
Please reply directly, as well as to this digest, if you would.
Dave Stoffel
seismo!gl-red!dave