home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.ai:5028 rec.arts.books:26367 misc.writing:4279 rec.arts.int-fiction:1284
- Path: sparky!uunet!inmos!fulcrum!bham!warwick!doc.ic.ac.uk!dapsun.lif.icnet.uk!not-for-mail
- From: js@montaigne.lif.icnet.uk (Jack Shirazi <js@biu.icnet.uk>)
- Newsgroups: comp.ai,rec.arts.books,misc.writing,rec.arts.int-fiction
- Subject: Re: Computer writes a book?
- Date: 26 Jan 1993 17:28:51 -0000
- Organization: Imperial Cancer Research Fund
- Lines: 21
- Message-ID: <1k3scjINN57j@montaigne.lif.icnet.uk>
- References: <1993Jan25.163029.1901@seas.smu.edu>
- NNTP-Posting-Host: montaigne.lif.icnet.uk
-
- In article <1993Jan25.163029.1901@seas.smu.edu> pedersen@seas.smu.edu (Ted Pedersen) writes:
- >>I saw the following in the Books column (by Olin Chism) of the Dallas
- >Morning News Sunday January 24, 1993. Copied without permission. It
- >concerns a computer program that allegedly wrote a book in the style
- >of Jacqueline Susann.
- >
- >I think this is a stunt. I don't really believe a computer program
- >wrote this book. However, if anyone has any information about this
- >book or the author that would give this some credibility I would like
- >to hear about it. Comment on the general idea would be interesting as
- >well. I don't think the idea itself is impossible, it just seems like
- >this case is a little far fetched.
-
- Markov chaining using a mixture of sentences and words, or possibly
- just with words, using a parser to get rid of any unlikely combinations.
- Probably added in the use of a style checker.
- --
- Jack js@biu.icnet.uk
-
- If you only have a hammer, you tend to see every problem as a nail.
- -- Maslow
-