Profile
Sentences
Vocabulary
Reviews
Lists
Favorites
Comments
Comments on marcelostockle's sentences
Wall messages
Logs
Audio
Transcriptions
Translate marcelostockle's sentences
really? there were just a couple typos.
Is it possible to link them back now?
can't seem to edit a couple of my sentences:
http://tatoeba.org/eng/sentences/show/1133031
http://tatoeba.org/eng/sentences/show/1136859
what have they got in common ? both have audio files attached
sry if obvious. been disconnected for very long
Gracias
how can I unadopt my sentences now?
Every time I log in I am taken to a certain page
http://tatoeba.org/fin/sentences/show/3204881
Welcome back.
Did you know the reason of my avatar being a snake is that with you, Shishir and a couple other members we get a significant part of the Chinese zodiac?
I think she means one doesn't need to know Turkish if your comments are trustworthy
+1
That is absolutely awesome!
I wish I had more time to participate
That is the right thing to do.
Tatoeba's Graph representation is useful in modelling inter- or intra-language subtle relations.
For example, you could also link all Bengali sentences you may use in the same context (formality, gender, number, etc.) to each other. And even with more languages.
The rule of thumb is "would this be a right thing to say in this precise context and meaning?". This way, even word-to-word matches aren't so significant.
Great news.
They are all important contributors.
I've noticed the changes in the German interface and I'm very pleased with them!
There one change that looks a bit awkward, though:
Sprachvoreinstellung für die Nutzung der Schaltflächen "Zurück", "Weiter" und "Zufällig"
Isn't that a bit too long? It skips a line and touches "Zeige Satz Nr.".
I can't really think of a proposal now.
I can't comment on the blog, but you may need to fix the date (year). It looks like old news
Felicitaciones Vivi! :D
Sentence adoption is not working
Para mí, este fue un laaaaaaaaaaaaaaaaaargo esfuerzo. Con un lenguaje como el alemán es muy fácil caer en la frustración en el proceso.
Solo te recomendaría leer harto, textos simples. Y muy importante, anotar cada palabra nueva en algún lugar (un cuaderno, una pizarra), ¡sin explicaciones!, para que las veas varias veces y te acuerdes por ti mismo.
+1
"...I should not bother about translations of translations, even if they happen to be identical with the sentence I am about to enter"
Yes.
And even if that creates complexity, duplicates, or any issues, it shouldn't be up to the CONTRIBUTOR to pay any attention.
"the waste of storage and CPU time, excessive complexity of graph structure and non-optimal search results"
I agree with all of that. I also work with the Tatoeba as graphs. And with all the complexity, all my algorithms suddenly stopped working. And that is detrimental.
But I don't think adapting the corpus to the computational needs is right, if not doing our utmost to adapt our computation to the corpus needs.
Well, I'll try to make it short:
Whenever there is a new duplicate, there is a chance other people try to translate it, even when the translation already exists.
So far, it sounds like a waste of time and effort. But it's not usually so, because when we think of a translation one more time, we may come to a new one that we didn't figure out the last time, enriching the corpus.
Is that a bad argument?
I don't think anyone should bother, given there is a de-duplication algorithm that "eventually" works.
It is quite an old discussion now. I think that time someone proved how they can become beneficial to the corpus. I could bring it back, but right now I really can't.