Can be used on PT-BR(Brazilian Protuguese)?

[Marciano De Andrade Junior]

The title says all.

[paulo paniago]

oi Marciano, o plover funciona com qualquer teclado qwerty, independente do layout.

Em segunda-feira, 4 de novembro de 2013 13h00min40s UTC-2, Marciano De Andrade Junior escreveu:

The title says all.

[Marciano De Andrade Junior]

Mas não é necessário um dicionário com as palavras? não tive muito tempo ainda para pesquisar sobre o plover, mas me interessei hoje e essa é a minha primeira necessidade, claro.

[paulo paniago]

Mas para usar varias teclas ao mesmo tempo, você vai precisar do SideWinder. (com o seu teclado, você pode usar o plover marcando uma opção chamada – arperggiate – aí você aperta as teclas que você quer, em qualquer ordem, e depois aperta o espaço).

[paulo paniago]

voce quer em português ou inglês? se for em português, voce vai ter q apagar o dicionário, e começar do zero. porque ele não existe ainda aqui. mas se for em inglês, o programa já vem com todo o vocabulário.

[Marciano De Andrade Junior]

hmmmmm certo, não tenho tempo para realizar a escrita de tal dicionário... gostaria em Português mesmo. muito obrigado paulo!

[Zack Brown]

I'm sure it's possible to make a Portuguese dictionary, but I'm not sure it could really be considered Plover at that point - it would be a whole new theory. But I'm sure the Plover software would support it.

2013/11/4 paulo paniago <ppani...@gmail.com>

eu posso te passar o meu, mas está cheio de erros, q fiz intencionalmente para não ter que digitar muito. e eu tb apenas comecei a fazer ele. tb tem q mudar o layout do plover antes da instalação. 

--
You received this message because you are subscribed to the Google Groups "Plover" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ploversteno...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
Zack Brown

[paulo paniago]

Hello Zachary, there is already a steno theory for Portuguese from stenograph. To use plover for that, it just requires minor changes in two files of plover before installation, so to change the layout. about a theory for plover in Portuguese, that  would be interesting. Maybe someday i′ll have my own product for that purpose. But, since I′m still learning, that might take some years.    

[paulo paniago]

I meant we have a theory of stenotype that was created by “Steno do Brazil” which is the sole dealer of stenograph in here. (Can’t believe I said our theory was from stenograph) – just needed to correct that point!

[Ricardo Pietrobon]

Paulo, I believe that the steno theory by "Steno do Brasil" is proprietary, and so a new theory would indeed be required. As Zach mentione, then this would no longer be Plover Theory but instead a brand new theory. That's a really tall order as the phonetics in Portuguese is completely different. I googled all over to try to find out how exactly Steno do Brasil did it, but apparently their theory is only taught through in-person courses in Sao Paulo, the courses not happening all that often.

As for the translation, from what I could gather the word "estenografia" in Brazil is used for shorthand, while "estenotipia" is what is being discussed in this group.

[paulo paniago]

Hello Ricardo, thanks for showing interest. Are you Brazilian? You are very well informed. I’m a little busy right now, but will answer you as soon as I can. 

[paulo paniago]

 

 Hello Ricardo, in case your considering to learn steno for typing in Portuguese and in English, realize that the layouts are different. That means that both theories are not directly applicable by the same training. so, before you start, unless your willing to learn to use two different layouts, you will have to decide, which will be your main typing.  For the other, you will probably have to do an adaptation, or simply use a qwerty keyboard.

[Ricardo Pietrobon]

all, Paulo and I exchanged a number of emails where he told me about his very impressive efforts coming up with a dictionary for Brazilian Portuguese. During this conversation I started wondering whether the creation of a steno dictionary like plover's is something that could benefit from other existing open source efforts or if it's something that necessarily has to be done exclusively by hand. Always bearing in mind that I am a total newbie to plover and stenography, here are some points and questions in no particular order:

1. Hunspell http://hunspell.sourceforge.net/ has a number of dictionaries in different languages containing not only long lists of words but also a standardized way of coding for affixes (sufixes, prefixes, verb endings, etc) - see http://goo.gl/Mhdcsw for the version applied to Sublime Text. the biomedical community has literally dozens of extensive, carefully curated taxonomies - mesh https://www.nlm.nih.gov/mesh/mbinfo.html is one out of many examples. Would lists of words like these be of any help or is the addition of words on the fly the most appropriate or only method? 

2. Would a "bag of words" ranked by frequency for a niche field (biomedical, education, IT, etc) be of help? I believe I saw Mirabai referring to an English word frequency list in one of her posts, but I didn't understand whether it did play a role in her dictionary or not

3. Would hunspell's pre-existing taxonomy of affixes help in any way in setting strokes for the plover dictionary? not saying they can't be modified later, but just to give a new language or niche dictionary a head start

4. Given that the plover theory sets a number of rules for how strokes can be coded, could some of the simpler rules be coded through regular expression so that (a) words from a pre-existing dictionary and following a specific pattern could be identified and then (b) the corresponding stroke would be automatically inserted in a dictionary. as always, the use case could be both a new language or a niche dictionary in English. I am sure these regular expressions would only cover a limited set of words before they start getting really hairy, but if feasible for the simpler cases it would allow new language or niche dictionaries to be jump started in a semi-automated way 

On Tue, Feb 11, 2014 at 2:36 PM, paulo paniago <ppani...@gmail.com> wrote:

 

 Hello Ricardo, in case your considering to learn steno for typing in Portuguese and in English, realize that the layouts are different. That means that both theories are not directly applicable by the same training. so, before you start, unless your willing to learn to use two different layouts, you will have to decide, which will be your main typing.  For the other, you will probably have to do an adaptation, or simply use a qwerty keyboard.

--

[psch...@gmail.com]

I have created a few scripts that I have used to convert a database of phonetic pronunciations of English words into dictionary entries.  I am sure that this technique could be adapted to other languages, provided there was a comprehensive source of pronunciation information (ideally broken up by syllables).

[Ricardo Pietrobon]

this is very interesting, how did it work for you? were you doing your conversion though regexp, sed/awk, lisp or something along those lines? very curious on how you matched the phonetics with plover theory or if you just gave it an initial go and then adjusted by hand later on

good news for Latin languages such as Portuguese or Italian is that the pronunciation tends to be more predictable than the one in English. also thinking that your method could perhaps be used straight out of the box for niche areas in English, where I guess the phonetics for technical words tends to be more predictable. the latter is just a a hunch, no hard evidence to back it up

one resource that immediately comes to mind in relation to this type of automatic conversion is wordnet http://en.wikipedia.org/wiki/WordNet since it is open source and contains phonetic information. The English wordnet contains around 115 thousand synsets, a synset being defined as group of concepts meaning the same thing. I am not sure exactly how many concepts/words this would correspond to, but it is massive. also available in multiple languages such as Portuguese https://github.com/arademaker/openWordnet-PT , although languages other than English tend to be more incipient

On Wed, Feb 12, 2014 at 3:59 AM, <psch...@gmail.com> wrote:

I have created a few scripts that I have used to convert a database of phonetic pronunciations of English words into dictionary entries.  I am sure that this technique could be adapted to other languages, provided there was a comprehensive source of pronunciation information (ideally broken up by syllables).

--

[psch...@gmail.com]

I think that I am pretty much done with English, as the source I used contained approximately 120,000 words (albeit without proper names), which is about 3x the vocabulary of the average speaker.  I turned each word into a sequence of strokes by mapping the initial consonants, median vowel, and final consonants of each syllable to parts of a stroke.  This is why using a source of phonetic information was important, because it removed most of the guessing that would have produced incorrect results.  I also added some simple rules for distinguishing between words with the same pronunciation based on their spelling (like cell and sell, for example).  The scripts themselves are actually in python.

[Mirabai Knight]

Hey, that's fantastic. Would you mind diffing it with the Plover
default dictionary to get a set of words that are both phonetic and
human-approved? That would be a great brief-and-misstroke-free starter
dictionary for use with Learn Plover!

[psch...@gmail.com]

I'm sorry, but the output doesn't correspond to the plover theory.  I could adapt it to produce a plover theory style dictionary, but I would need some guidance.  For example, how to deal automatically with conflicts such as surf / serf (and representing the rf ending without conflicts), how to represent the lk ending without conflicts, and how to distinguish ft / st without creating conflicts.  And that's just for starters.  Because I didn't know the plover approved way to deal with these problems, I went to another theory.

[Hesky Fisher]

Would you mind sharing your code for generating the dictionary?

--

[psch...@gmail.com]

Sure, but it needs to be gotten into shape to be shared.  Currently it is actually a sequence of scripts each transforming the file for the next, which is a product of how they were gradually developed.  I'll polish them up into a single program as soon as I get done making the Stentura work with StenoLite

[psch...@gmail.com]

Ok, I put the files that I can share, makedictionary.py and map.txt in the StenoLite repository.  However, I don't have the rights to distribute the raw phonetic data itself or the word frequency data.  So you can either find your own sources or you can give me the modified script and map file back and I can run them with my data sources.

On Wednesday, February 12, 2014 5:01:11 PM UTC-8, Hesky Fisher wrote: