[prev in list] [next in list] [prev in thread] [next in thread] 

List:       aspell-user
Subject:    Re: [Aspell-user] probabilities
From:       Lars Aronsson <lars () aronsson ! se>
Date:       2003-02-10 13:10:06
Message-ID: Pine.LNX.4.44.0302101403320.9410-100000 () bryant ! aronsson ! se
[Download RAW message or body]

Kevin Atkinson wrote:

> > Shouldn't probability or frequency be weighed in when Aspell
> > determines which correct word is most likely?
>
> I have considered doing that for a long time.  It would defiantly help.  I
> have also considered doing so on a per document basis.  I just never got
> around to doing so.

It is of course related to the question of where to find such data.
Today's spelling dictionaries don't contain information on which
words are more or less frequent, so this information base would have
to be built.  Perhaps it would suffice to indicate which words are (1)
among the 1000 most frequent words in a language, (2) very rare words
that maybe only appear once in a million words, e.g. names of
fictional persons or places, (3) all the rest.


-- 
  Lars Aronsson (lars@aronsson.se)
  Aronsson Datateknik - http://aronsson.se/




[prev in list] [next in list] [prev in thread] [next in thread]