it`s great to have an own mailinglist. I what you to know that I like to join
and share some ideas for wikitionary soon. I came to wikipedia to find an
explaination for words that I haven`t found with ispell... more about this
visit (still only in German): http://de.wikipedia.org/wiki/
At the moment I try to get a 25.000 word-list (catchwords) from a geology
bibliothek for a use with GNU-FDL. Beside of wikitonary, I have at he moment
a disput on the German Wikipedia Mailinglist - about a better reserch of
catchwords and an algorithmen how to give names to articels with less as
posible subjectivity. *g*
Because of making my mind about more systematic I tried to scan (manualy) all
wetterforcast region of an german wetterreport:
Theses regions are not realy a good example for translations, but it should be
an example to scan the text of national radiostation for words of one field.
With such kontext is it more easy to translate and more usefull for learning
vocabulary. The sum of all fields could build the wiktionary, and one link of
rain could show a list of meterologic words (rain, snow, storm...)
I think the software should support the research of the words and the wictors
(viktors (what is the name for wikitonary people ?)) should work as
systematic as posible and build wordlists for seperate sujects bevore to link
them to another language. To build software for Wikitonary is definitive more
dificult than for the wikipedia, and beside of software-tools giving
software-skills to the wictors should have the same priority.
So with "quick and dirty" we won`t have succsess, and we should take time for
- how to get lists of words (scaning ebooks,news, catchwords form bibliotheks)
- how to split these list in fields
- how to work with them
When we find all ways how to find words systematicaly, for every way we should
creat a dummy wordlist to know how to handel these.
Back to the 25.000, I want to add statisticaly informations to these words,
how many books, news, articals, dissertation, webpages, newsgroups-posting
this words are used.
Greetings from Aachen,
PS: I study civil ingeneering ;)
This wiktionary mailing list was created to discuss some possible changes,
implementations, and improvements to the current Wiktionary. The idea is to
transform the current freeform wiki-dictionary into a more formal structured
format. That way things can be more easily searched, linked, and managed.
The debate is how simple or complex should the system be. If it is too complex
no one will add to it, if it is too simple, then the conversion from wiki-system
to the new format was in vain. We also don't want to reinvent the wheel, there
are several other dictionary format available. One for example is the DICT
protocol. The Dictionary Server Protocol (DICT) is a TCP transaction based
query/response protocol that allows a client to access dictionary definitions
from a set of natural language dictionary databases. RFC 2229 has more
information. This might be an intersting format to follow, or maybe not?
The system should be fairly simple and expanded later. The major points to be
broken down are: Words (terms, various prononciations, language), Definitions
(part of speach, definition, example), synonyms, antonyms, etymologies,
translations? Each would be it's own table keyed by the word. This is a very
simple system, and it is a good point to start a discussion.
The other table i would suggest is some sort of 'See Also'. This would be used
for things like mutiple spellings, or plurals. So if i enter 'colour' and color
is already in the dictionary, i can add a definition to 'colour' as "The british
spelling of color"... SEE ALSO: color. The same would work for plurals (goose,
geese). This doesn't force any weird root word constrains, and allows for some
miskates by people NOT check for alternate spellings first, but is easily
fixable with a 'see also' link.
that's my two cents, any suggestions?
 - http://www.dict.org/bin/Dict
 - http://www.dict.org/rfc2229.txt