Zdenek Broz of the dicts.info project has done some interesting
analysis of a recent
en.wiktionary.org dump:
http://wiki.webz.cz/wikt.png
This bar chart shows the number of translations within articles that
have "Translations" group headings at all (note that most
en.wiktionary.org pages don't have any). Accordingly, there are only
21031 translations to English, followed by 14341 to French and 13004
to German.
While
en.wiktionary.org has nominally 313537 pages, the vast majority
of these are tiny stubs with short definitions, but no translations.
For comparison, OmegaWiki (formerly WiktionaryZ), which is still
pre-alpha software, currently has:
175990 expressions ("pages" in Wiktionary lingo)
11948 DefinedMeanings, i.e. concepts (comparable to the total number
of separate definitions; Wiktionary has much more of these)
117 active languages
10964 English expressions
11277 German expressions
9342 French expressions
etc.
This isn't an entirely fair comparison, since Wiktionary, due to its
architecture, is split into many language databases. One could
potentially derive more translations by parsing and combining data
from the different editions. Of course, the lack of a machine readable
format makes that difficult. Still, it can be said that Wiktionary, as
of now, isn't particularly useful as a _translating_ dictionary. It's
very useful as a _defining_ dictionary, though (arguably a very
similar application to Wikipedia with much shorter texts).
On the other hand, the architecture of OmegaWiki seems to be validated
by its steady growth. I'm fairly happy with our backend, but our
frontend sucks; data entry is still much too cumbersome. This will be
our big challenge for next year -- taking the project to a final,
feature complete release that is user friendly. But I think we're on
the right track.
--
Peace & Love,
Erik
DISCLAIMER: This message does not represent an official position of
the Wikimedia Foundation or its Board of Trustees.
