coregraf-developer Mailing List for Spellchecker : next generation

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

Hi,

the "orthomaker.cpp" part of the program was designed to be the  
combination generator of the application
the "exavar.cpp" was designed to be the variant generator
the "notevar.cpp" was designed to evaluate variant :

      in fact only the grammatical parsing was setup to work  
correctly but other value can be used.
      other value give us two problem :

               how to get statistics ?
                      for example how to have statistics about the  
verb "être" for each form "suis","es","est" ... ( and for each user  
age ...)
                                           "est" is frequent "fûmes"  
is rare

                      how to use statistic ? small grammatical word  
was over valued / single dictionary words

              how to balance each value ?
                         I have noticed that grammatical parsing is  
more effective in verb group with strong grammatical structure ( e.g.  
"je ne l'ai pas vu")
                              statistics may be more effective for  
long name noun group ( e.g "un extraordinaire éléphant")

         there's also some sketch of other possibilty in the application
                      - limitation of sens
                      - valued variant when  the same tags are  
matched in the sentence ( in the text ?)

                      I have also an idea for the future of  
calculation of the distance between words in the dictionary.
                                need to build link between words   
( "who can have what ?" "Who can be what ?" "Who can do what ?)

"orthomaker" and "exavar" was not the corrector , it was  
"l'imagination" of the programm , they try to open all the possibility.
the other part try to closed them as well as possible.

Jean-Pierre

Le 29 janv. 11 à 20:08, Andy Guy a écrit :

> Hi
>
> Before I continue, I realise that I didn't make it clear that I wasn't
> talking about the overall application, but just the component (or  
> module, in
> the future) that does the work; the component that the application  
> sends the
> letters/syllables/words to, and which provides the solution. A lot  
> of what I
> have been working on is involved with separating this component out  
> from
> orthoconsole so it can be reused with other (including GUI)  
> applications.
>
> What I was picking up on was the fact that the "spell-checker"  
> already does
> more than check for orthographical errors.
>
> As I understand it, orthography is specifically about writing; it is
> spelling, hyphenation, capitalization, word breaks, emphasis, and
> punctuation. It does not include the actual rules of grammar, which  
> are used
> both when you write and speak.
>
> Your code already uses grammatical analysis to work out if a given  
> word
> order is permitted; it might be possible to feed this information  
> back to
> the user, to let them know when what they're writing does not make
> grammatical sense. But this is a grammatical error, not an  
> orthographical
> one.
>
> For example: "Le petit noir chien." is correct in an orthographical  
> sense,
> as the spelling, capitals, and punctuation are all correct. You can  
> even see
> what they mean.  But it is incorrect according to the rules of French
> grammar.
>
> While reading round the subject I have seen people complaining  
> about the
> inability of others to distinguish grammar from orthography. It  
> also appears
> that there is some debate about the exact boundary between them. So I
> thought that the safest thing to do would be to avoid specifying  
> exactly
> what was being corrected.
>
> But, as I said to start with, this is the name of the "corrector"  
> module;
> not the overall application.
>
> Regards, Andy
>
> -----Original Message-----
> From: Jean-Pierre Soto [mailto:jea...@wa...]
> Sent: 29 January 2011 16:08
> To: Andy Guy
> Subject: Re: [Coregraf-developer] Spell-checker, Corrector,  
> Orthomaker, ...
>
> Hi,
>
> I 'm not sure to understand what you mean.
>
> The more open term is "Coregraf". The basic meaning is "Correcteur
> orthographique". But over this meaning, the dance term "corégraphe"
> show that we must call all the possibility ( phonetic, grammatical  
> parsing
> ,dictionary , text-to-speech ) to work together to help user.
>
> As the child do when they learn to read, they build bridge between
> separate working task and separate part of the body
> (memory,eyes,hand ,ear ... )
>
> Jean-Pierre
>
> Le 29 janv. 11 à 16:40, Andy Guy a écrit :
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 9.0.872 / Virus Database: 271.1.1/3409 - Release Date:  
> 01/29/11
> 07:37:00
>

2009	Jan	Feb (2)	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2010	Jan	Feb	Mar (1)	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2011	Jan (7)	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec