Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!mcgill-vision!snorkelwacker!usc!zaphod.mps.ohio-state.edu!wuarchive!decwrl!elroy.jpl.nasa.gov!lll-winken!cert!netnews.upenn.edu!eniac.seas.upenn.edu!jeffe
From: jeffe@eniac.seas.upenn.edu (George Jefferson )
Newsgroups: comp.sys.mac.apps
Subject: Re: How do so many words fit in a dictionary/thesaurus?
Message-ID: <29477@netnews.upenn.edu>
Date: 12 Sep 90 21:50:30 GMT
References: <3374@stl.stc.co.uk>
Sender: news@netnews.upenn.edu
Reply-To: jeffe@eniac.seas.upenn.edu (George Jefferson )
Organization: University of Pennsylvania
Lines: 15

In article <3374@stl.stc.co.uk> pac@stl.stc.co.uk () writes:
>Can anyone give me a reference to the compression techniques used by the
>dictionary/thesaurus applications and DAs.  Taking Microlytics claim about
>Word Finder (220,000 words) it works out at about 1.6 bytes per word.

I suspect that the '220,000' words counts plurals, and other word
variants as seperate 'words'.  A tremendous 'compression' is
achieved by recognising the (reasonably) simple rules
for adding aprporiate suffixes and prefixes.  Eace base word
need only cary around a couple of bytes to indicate which modifiers
are allowed.

-- just a hunch

george