हमारा डेटा विभिन्न क्रिएटिव कॉमन्स लाइसेंस के तहत जारी किया गया है।अधिक जानकारी
यदि आप इस सामग्री से प्यार करते हैं, तो कृपया दान पर विचार करें।
वाल (7,136 धागे)
नुस्ख़े
सवाल पूछने से पहले, फ़ॅक जरूर पढ़ें।
हमारा उद्देश्य सभ्य चर्चाओं के लिए एक स्वस्थ वातावरण बनाए रखना है। कृपया हमारे बुरे व्यवहार के विरुद्ध नियम पढ़ें।
mraz
4 दिनों पहले
mraz
9 दिनों पहले
mraz
9 दिनों पहले
Dovud123
9 दिनों पहले
sharptoothed
9 दिनों पहले
frpzzd
17 दिनों पहले
hecko
17 दिनों पहले
frpzzd
18 दिनों पहले
araneo
18 दिनों पहले
gillux
18 दिनों पहले
Concerning copyrighted sentences extracted from textbook
http://tatoeba.org/eng/sentence...comment-165797
I deleted the sentence before seeing the comment so I answer here, and anyway I think it may be informational for other people
Here actually I had a high doubt that the sentence was natural, though it was a very simple sentence, I never heard such a construction. So I've googled it
https://www.google.com/#hl=en&q..."&oq="球被他踢进去了"
3 results (so very very uncommon sentence, especially as two links are non native, and the third one is a translation of an article in English, and knowing Chinese people quite well, it's likely to be a simple copy paste of google translate or so)
Moreover the first link is a "website-ified" version of a textbook I know as my students here use it too to learn Chinese. If you take a look at the other sentences of this user, they are exactly the sentences extracted from the same exercises, and for the others they also give the same kind of results (very few link , less than 10, and pointing to this book, sometimes it require a little more investigation as it's a "put the character in the right order")
So all of this made me pretty sure that the "block" of sentences he added come from that book.
Of course you can't copyright "he's eating an apple", but adding exactly the same sentences in "block" then start to fall under things that can be protected by copyright (otherwise any books is only a set of words that everyone can use)
After I know that people who have been here for a long time know my position on that subject, but I will still precise it for the new users. I also find this situation stupid, and I would also like being in a world where "open culture" project such a wikipedia, tatoeba, universal subtitles, global voices etc. can get rid of this copyright stuff, but unfortunately it's not and some of these project (I mainly think about the handedict project) who didn't pay attention to this needed to be closed during some months (the time to "clean") because some users started to enter too much copyrighted data from other sources
So here I really don't want to take the risk to be sued, even if it's very very low, but the impact on the reputation of Tatoeba, the time it would take to check the database etc. is too huge, so I prefer to be quite strict on this.
After I don't blame the users who thought in a first time that it was not a big deal to do this. I just hope they understand after why I must to do it.
Please add a tool to check the duplication of sentences.
Can you forbid a user to post 2 exactly same sentences ?
Think of this situation.
This month, I post 1000 sentences. Next month, I post 500 new sentence and 500 old sentences (which I have posted before).
they are deduplicated regularly and their translations, comments and tags are merged.
To know that a sentence is a duplicate, you need to type it anyway...
What's usually done about users who have sentences to be revised, but don't seem to be active any more? Is it possible to force a de-adoption?
Moderators, aka corpus maintainers, are supposed to fix those sentences after a certain period of time has elapsed.
Hello esperanto users. I was wondering what's your method to write esperanto special diacritics, since I will probably start studying this language soon.
You might want to check out PhraseExpress. I have it set up that "c#" gets converted automatically to "ç" etc.
http://www.phraseexpress.com/
Seems neat, but I dislike the idea of something modifying my input, even when it's about niceties that I would rarely find myself typing at all.
I guess I'll stay with this not-completely functional layout for now, thanks everyone for your help anyway.
Thanks, although I can't seem to find that layout on the ones I can add to my keyboard layouts. Perhaps I'll Google tomorrow.
.-.-.
Whops, I forgot to specify that I'm under Windows, and although I used to love that freedom Linux gave me when typing, I can't nor want to switch to it again.
I tried to build a keyboard layout via Microsoft's own Keyboard Layout Creator but it turn out the diacritics of esperanto (ˆ and ˘), which I want to use to make dead keys, are not supported in the italian mapping and I know of no language that may have them which I could actually use.
This page has a few options: http://en.wikipedia.org/wiki/Es...rosoft_Windows
Thanks, but neither of them seem to work or be good enough for one reason or the other.
Somehow I managed to get my original system to work but it sill fails in some non unicode fields and programs causing for example ĉ to be displayed as cˆ, and I don't know what to do to solve that or if there even is a solution.
Thanks anyway and if there is any other esperanto enthusiast or speaker using Windows, I'm all ears.
btw, there exists an all-time virus in windows, very common, which won't let you write "two-keys-characters" like èéëŭŝ
have you made an antivirus scan lately?
if it is that virus, then a simple scan always mends the problem
Luckily no, I CAN use dead keys, but they do not work in non-unicode environments because of the language map or something.
Use a German mapping. Don't know why it doesn't work for Italian, but I extended the German keyboard mapping to deliver all kinds of accentuated letters, so it does work with this one.
2 IMPORTANT SUGGESTIONS to the HOST:
For translators, it would be great to be able to select which language or which combination of languages to be excluded, so that he can find only the sentences that haven't been translated to his target language.
For readers:
It would be very useful for them to be able to look up only the sentences that combine, at once, translations of those languages they want to see together.
You can lookup sentences in any language you want : http://tatoeba.org/fre/sentence...ne/indifferent
see picklists on the right column...
Yes, but you can't choose random combinations. For example someone who wants only sentences that are translated to say German, French and Spanish at the same time, it would be very useful to allow any number of selected translations. Another thing is, when a t
contributor wants only those sentences that are not translated to his target language (so as to be faster) he cannot choose the languages to exclude. One of us could be, say, looking for sentences which have both Japanese and Russian translations and lacking Turkish and English translations, it would be nice to offer that flexibility...
if you need some kind of specialized query, you can send me an e-mail with your query
find more information in my profile
note: I think developing and offering query flexibility is not part of the aim of the project
Well, I understand. But, I came up with that in order to be able make faster contributions. I am having to spend a lot of time trying to find the sentence that is not translated into my target languages. It might be that others are facing the same problem. It would be a part of the aim of the project for people to make more contributions.
I think being able to select sentences in one given language that are not translated to another is already good enough...
I agree, but that would certainly overload the server.
TATOEBA can not search for the right word in Unicode.
try to search for the example of this word "vải" (in vietnamese, it means fabric)
you will have sentences with "vai" "vài" ...
Please check and fix this.
yes I understand your request but what if a vietnamese or soemone with a keyboard without possibilities to input these diacritic want to search for something. I do understand your request as even in my own mother tongue (french) it can be problemati, but in an other way I do understand that many people in either a educational or professional context (i.e on computers on which they cannot install their own input method), can not enter these diacritics (even on a cellphone their own themselves)
A good solution could be the one adopted by Google search engine: a character with diacritics finds the one with diacritics, while the same character without diacritics finds them all.
I understand Trieuho's problem. If I, for example, wanted to search for italian sentencences containing "perchè", which has the wrong accent, I couldn't do that.
In the end, the only languages that aren't affected by this problem are those who don't use diacritics, and those are really a ristrected few. Tatoeba doesn't focus on those language, it consideres all the languages as equal, so they should all have the same possibilities.
"A good solution could be the one adopted by Google search engine: a character with diacritics finds the one with diacritics, while the same character without diacritics finds them all."
-- Shadd
YES ... That's it ... That's a best solution and that's also what I mean. (Thanks Shadd, English is not my mother language so it's hard for me to explain.)
Ok, I will try to see if that's possible to have this with the current search engine we're using (Sphinx, for those interested, I'll be glad actually if someone already have the solution and want to share it with us)
How about chinese ? When you learn chinese, you can not say "I don't have an input method, how can I search for sentences in Chinese characters ?"
"Vải" & "Vài" & "Vai"
Those are 3 different words in Vietnamese. If you don't have an input method software, you have to accept to search with the word "Vai" (no mark).
So ... as what you say, Tatoeba is lack of one more function. "Input Method System".
You can enter this site to type Chinese.
http://userwww.service.emory.edu/~zxie3/ime.html
You can enter this site to type Vietnamese.
http://just.nicepeople.free.fr/...ese-Typing.htm
I am not a software programmer but I think you can intergrate the input method into this site.
By the way, I can use my cellphone to enter Chinese, Vietnamese, French ... (Samsung GalaxyY)
My opinion is to make Tatoeba to become a best example library.
*** mezaĝo de la Tatoebaj kontribuantoj por ĉiu lando ***
Bonvolu vidi ĉi sube la liston de la mezaĝoj de la Tatoebaj kontribuantoj por ĉiu lando kun pli ol 3 konfesintaj kontribuantoj
lando kontribuantoj mezaĝo
------------------
portugalo 3 60,00
svedujo 5 46,50
belgujo 13 45,82
hungario 13 41,50
vjetnamio 4 40,67
colombio 3 39,50
germanujo 54 37,94
irano 6 37,60
japanio 13 36,20
aŭstralio 5 35,00
singapuro 3 35,00
francujo 59 33,42
argentinio 11 32,56
danujo 6 32,25
brazilo 39 31,97
italujo 18 31,82
ukrainio 12 30,78
hispanujo 44 29,96
polujo 31 29,53
aŭstrio 6 29,50
rusujo 39 29,39
nederlando 14 29,00
britujo 18 28,50
finlando 13 28,40
kroatio 3 27,67
rumanio 4 27,50
ĉilio 7 27,40
kanado 13 27,20
usono 38 27,00
turkio 10 27,00
meksikio 9 26,83
indonezio 6 25,50
tajvano 4 25,25
ĉinujo 10 23,80
norvegio 3 23,00
hong-kong 3 17,00
barato 3 17,00
israelo 8 13,00
Is there a way to distinguish the original sentence in a group?
Apologies if this has been asked and answered somewhere else. I couldn't find it.
I have a script where I can get most a user's original sentences, although it also mixes original sentences adopted by other users.
If you want someone's in particular, I could send them to you.
Gracias, Marcelo. I dont' think I need that, but It's good to know. I was just trying to avoid a faux pas in the future and now I know what to do. :)
OK,
if you ever need to make such a query or any other personalized query, you can find more information in my profile : )
cheers
Podrias hacer una estadistica para saber desde que y a que idomas se traduce el esperanto.
Tal vez tambien seria interesante tener la misma estadistica del ingles para comparar.
Supongo que habra mas interrelacion en lenguas orientales y esperanto que con el ingles. Aunque no lo ser.
Estatisticts about translation between ¿among? esperanto and other languajes. From esperanto and to esperanto.
Many thanks Alexmarcelo :)
OK, there you go:
Esperanto:
31,66% translated to 'eng' (19,46% of 'eng')
34,82% translated to 'fra' (41,70% of 'fra')
32,86% translated to 'deu' (43,57% of 'deu')
12,92% translated to 'spa' (17,70% of 'spa')
9,47% translated to 'por' (21,59% of 'por')
4,46% translated to 'ita' (11,53% of 'ita')
6,72% translated to 'rus' (23,76% of 'rus')
5,11% translated to 'pes' (78,53% of 'pes')
2,41% translated to 'dan' (62,85% of 'dan')
Spanish:
70,95% translated to 'eng' (31,57% of 'eng')
13,83% translated to 'jpn' (9,10% of 'jpn')
17,71% translated to 'epo' (12,93% of 'epo')
15,01% translated to 'fra' (12,06% of 'fra')
19,66% translated to 'deu' (18,80% of 'deu')
26,91% translated to 'por' (41,32% of 'por')
5,20% translated to 'ita' (9,08% of 'ita')
I'm only displaying the biggest results (more than 2% translated to ...)
si necesitas algo más, escríbeme
I hold permanent cross-translations statistics here, which I update everytime there is a significant increase in the number of sentences in any language https://docs.google.com/spreads...JNl92Wnc#gid=0
The sheet has different other tabs of different interesting statistics on contributors.
Interesting. I go to see it. :)
Que interesante, Gracias :)
Que igualadas estan las traduccion de Epo a eng, fre y deu.
Hay un porcentaje muy alto de las frases en Fra y Deu de origen esperanto.
Y el persa y el danes extremadamete alto, aunque sea un porcentaje pequeño de las traducciones desde esperanto.
Tal vez seria interesante una estadistica de los idiomas con el porcentaje más alto de sus frases traducidas desde el esperanto.
Indicaría la importacia del esperanto para ampliar el corpus de eso idiomas.
Me ha sorprendido no encontrarme ningun idioma oriental entre las traduccion desde el esperanto.
¿Podrias hacer una estadistica con los idiomas japones, chino y coreano?
Las traducciones del español al japones, epo, fra y deu estan muy igualadas.
Un tercio de las frases en ingles tienen traduccion al castellano. No me esperaba un porcentaje tan alto.
Gracias ha sido muy interesante :)
Sí, ningún problema, la hago apenas llegue a mi casa, y la publico aquí.
En general yo veo eso por usuario, por ejemplo,
"de o a qué idiomas están traducidas las frases de [usuario]"
o cosas así.
las del inglés pueden ser un poco desbordante, ya que prácticamente todos los idiomas están traducidos en parte al inglés. Así que pondré las del esperanto y el español nada más.
por cierto, soy Marcelo Alejandro ¬¬
Gracias :)
Perdona, siempre os confundo a alexmarcelo y a ti. :P
Menos mal que no os habeis puesto marceloalex y alexmarcelo...
Usually (there are historical exception), the oldest sentence has the lowest number.
However, this has no importance whatsoever, since translating is always betraying, and hundreds of translations and retranslations may have taken place from one given end of a translation tree, so that looking for the original is nearly impossible, but also vain and pointless: Tatoeba operates in trees and not in blocks of sentences. Each sentence might have dozens of variants of translations in one single language and theoretically thousands in the hundred+ language that we now feature.
You should thus only focus on translating the sentence at the top of the pile when you push on the translate button. That is the one you are translating, and nothing else.
The "original", provided you ever find it, should have no bearing whatsoever on your translation.
For one given sentence, the column on the right indicates all the translation links and when they were created (or deleted).
Thank you. I asked because I found two sentences linked together where one meant the exact opposite of the other. Now I know that I can ask for the sentences to be unlinked instead of leaving a comment.
Still figuring things out around here. :)
Pay attention to the colours of links:
Green links are DIRECT translations.
Grey links are INDIRECT translations (translations of translations)
It might well be that indirect translations sometimes contradict each other in a legitimate way. Languages do not match each other word for word, And I would not be surprised that "Yes" is indirectly translated to "No" through certain languages, somehow, like the French « terrible » can mean the exact opposite of the English "terrible" in certain instances...
Actually at a moment M, the one at the top is the "main" sentence from which you see the sentence directly connected to it (bright green) and indirectly (means it's a translation of one of the bright green translation) connected sentence (pale green)
So after if at one moment you two sentence directly connected that both taken separatly means something, then they're not one "incorrect" that should be changed according to one "original", but more two "group" which need to be split and a new sentence (actually two), to be added.
Has audio and icon.
If you search in "has audio" tag, You see sentences, but not the speeaker icon.
You can click on sentence and after click speeker in the new page. But click in firt page is easier.
An impovement for de future :)
Thanks sysko and Trang :)
browse -> browse by audio, ;-)
the audio tag has been created I think for those who do automatic treatment on it, as that's the only to get this information exported for the moment.
I should see this. :-/
Many thanks.
Other question. Can i search sentences with a word a sound example?
I try this hacking a don't work..
http://tatoeba.org/spa/sentence...earch?query=do
I hope this be a inteligent cuestion..
Thanks in advance :)
A school in Germany planning a day dedicated to counter extremism would like this sentence translated into as manny languages as possible.
1545891
Germana lernejo petas tradukon de cxi tiu frazo en laux eble plej multajn lingvojn.