»

Astuces

Ici vous pouvez poser des questions d'ordre général comme par exemple comment utiliser Tatoeba, rapporter des bogues ou des comportements déplacés, ou simplement socialiser avec le reste de la communauté.

Avant de poser une question, soyez sûr que vous ayez lu les FAQ.

Derniers messages

Vous pouvez écrire dans la langue de votre choix. Dans Tatoeba, toutes les langues sont égales.

Mur (67 sujets)

<<
<
12345 >>>
  1. slavturk = boracasli...
      • Thanks, sacredceltic. We'll take care of it.
  2. Pessoal, eu gostaria de saber se é possível fazer uma busca no site das frases em inglês com áudio e sem tradução para o português mas em uma ordem aleatória. Eu estou traduzindo estas frases, mas estou ficando um tanto quanto entediado porque as frases sequenciais são sempre relacionadas ao mesmo assunto. Por exemplo, das últimas 30 frases que eu traduzi, todas estão relacionadas com a língua francesa. Ou seja, como diz uma das frases traduzidas: "Eu já estou cheio da língua francesa".

    A busca que eu estou utilizando é a seguinte:

    http://tatoeba.org/pt_BR/senten...nly-with-audio
    • Translation to English:

      Hello everyone, I'd like to know if it's possible to make a random search of english sentences with audio that are not yet translated to Portuguese. I'm using the search above, but I'm getting pretty bored, because the last 30 sentences I've translated were about French:

      http://tatoeba.org/pt_BR/senten...nly-with-audio
      • above = below
      • French is as good a topic for a grammatical sentence as anything else...
        • I agree with you but I wanted to learn other words too. The last sentences I've translated have no new words.
      • Marcos,

        I also frequently want to visit a random or pseudorandom page within a set of pages. Here's a trick that someone taught me. It requires you to pretend to be a random number generator yourself.

        Go to the URL you mentioned, then to a later page in the set (for example, by clicking on the square marked "2" or the square with the ">" symbol). This will give you a URL like this:

        http://tatoeba.org/eng/sentence...h-audio/page:2

        (You can bookmark this page for future use if you want.)
        Now, generate a pseudorandom number in your head. Let's say it's 15. Replace the "2" at the end of the URL with "15" and you can jump to this "random" page.

        It would be nice to have this done automatically, since humans are surprisingly bad random number generators, but at least it's a start.
        • Thank you, Alan. I think this will help me a lot.
    • Traduko al esperanto:

      Saluton al ĉiuj,

      Mi scivolas ĉu eblas fari hazardan serĉon por trovi anglajn frazojn kun sono, kiuj ankoraŭ ne estas tradukitaj al la portugala. Nun, mi uzas la serĉon ĉi-sube, sed ĝi estas enuiga ĉar multaj frazoj estas samrelataj: la lastaj 30 frazoj kiuj mi tradukis, ĉiuj el ili, pritraktis pri la franca.

      http://tatoeba.org/pt_BR/senten...nly-with-audio
      • Saluton, Marco,
        mi supozas, ke simple ne ĉiam ĉiuj revoj estas realigeblaj.
        Tamen jen tre modesta konsilo:
        Unue uzu la ligon, kiun vi donis en via supra mesaĝo
        http://tatoeba.org/epo/sentence...nly-with-audio
        Poste uzu la butonojn super la ekzemploj
        << < 12345 > >>
        por navigi al alia paĝo, ekzemple du.
        Nun rigardu la adreson en la adresofenestro de via retumilo (browser).
        http://tatoeba.org/epo/sentence...h-audio/page:2
        Ĉe la fino vi vidas nun (ne surprize ☺) la nombron du.
        Vi povas nun marki tiun nombron kaj entajpi alian nombron. Por doni ekzemplon mi aldonis nulon.
        http://tatoeba.org/epo/sentence...-audio/page:20
        Tiel vi povas salti trans laŭplaĉe granda nombro da paĝoj, kaj tiel vi pli facile trovas frazojn, kiuj interesas vin. En la sama maniero, vi povas reveni tien, kie vi hieraŭ finis vian tradukadon por daŭrigi ĝin hodiaŭ. Por iri unu aŭ du paĝojn (aŭ al la komenco aŭ fino) kompreneble sufiĉas la menciitaj butonoj << < 12345 > >>.
        • Dankon, Alexander. Tiu solvo estas sufiĉe bona al mi.
          • Ĵus mi rimarkis, ke Alan donis al vi fakte la saman konsilon. Nu, ne gravas. Prefere ricevi du konsilojn ol nul, ĉu ne. ☺
            • And this way, we have an answer in two languages!
            • Ambaŭ vi pravas.
            • Both of you are right.
    • Salute,

      Io volerea saper como facer un recerca,, al hasardo, de phrases anglese con sono ancora non traducite al portugese. Io usa le recerca a basso, ma illa es enoiose pro que omne phrases es simile. Le ultime 30 phrases que io ha traducite contineva similemente le parola "French":

      http://tatoeba.org/pt_BR/senten...nly-with-audio
  3. Je suggère un affinement de l'algorithme de reconnaissance de langue.
    Ma phrase : « J'ignore exactement où. » a curieusement été identifiée comme de l'anglais.
    Il faudrait à mon avis, en sus des séquences de (3 ou 4 ?) caractères, prendre en compte les jeux de caractères eux-mêmes, puisque la lettre « ù » ne peut pas figurer en anglais.
    Mais aussi exclure les combinaisons impossibles. "J'" est impossible en anglais. Et je pense qu'une apostrophe devant un "i" l'est aussi...à confirmer...
    • +1

      Ça m'est arrivé parfois d'écrire des questions en espagnol avec "¿" au début qui ont été identifiées comme anglais ou français...
    • En fait il faudrait aussi que je remette à jour les données du serveur en production, il me semble qu'il est encore au statistiques d'il y a un an.

      Le but est aussi d'avoir le code le plus simple possible / adaptable, donc je pense plutôt que dans le calcul du score, pour les langues ayant suffisamment de données, considéré que tous les "N-grams" possible y sont présent , et impliquer un score negatif si un n-gram d'une phrase a detecter n'est pas présent.
      • > En fait il faudrait aussi que je remette à jour les données du serveur en production, il me semble qu'il est encore au statistiques d'il y a un an.

        C'est très vilain...enfin je ne t'en veux pas trop parce que je sais combien tu es occupé avec tout ça et je t'en remercie.

        >Le but est aussi d'avoir le code le plus simple possible / adaptable, donc je pense plutôt que dans le calcul du score, pour les langues ayant suffisamment de données, considéré que tous les "N-grams" possible y sont présent , et impliquer un score negatif si un n-gram d'une phrase a detecter n'est pas présent.

        Mais tes N-grams font-ils la différence entre un « u » et un « ù » ou un « ? » et un "¿" ?

        Amastan se plaint que le berbère est confondu avec l'espagnol et le finnois alors que le berbère comporte la lettre spécifique "ɣ"...
      • et j'ajoute que je suis très fervent des algorithmes les plus simples et que ton idée de statistiques de N-grams est brillante. Mais on peut ajuster...
  4. Why is it that unsubscribed users cannot see the list of sentences of a particular user?
    http://tatoeba.org/jpn/sentences/of_user/tommy_san
    This is pretty inconvenient when I want to show my friends what I'm doing here.
  5. Hello, everyone.
    I wrote a script to transform Tatoeba light view [tatoeba.org/s/s/400] into a quiz:
    http://userscripts.org/scripts/show/170945

    Details of usage are in the description.

    Could anyone tell me if everything is running OK?
    • Hola, Marcelo,

      Sí funciona, y me encanta. El único problema es que puede ser difícil encontrar una oración en el idioma que querés para traducirla a otro idioma específico. Es posible crear un filtro o algo del tipo? Creo que tu "motor especial de búsqueda todo-propósito" podría ayudar. Por ejemplo, si quiero una relación de oraciones en portugués para practicar traduciéndolas al español, al francés y al inglés, necesitaría un filtro que me diera solamente las oraciones en portugués que tienen todas esas traducciones. Qué decís vos?

      Otra cosa interesante sería que viera solamente los idiomas que me interesan, sin tener que cambiar las preferencias en mi perfil.

      Gracias por el trabajo.
      • Todas estas extensiones son posibles y precisamente a eso apunta esta idea. Esta es solo la primera pieza (la que más me costó concretar) de la herramienta que de seguro tienes en mente.

        Ahora estoy buscando la mejor forma de integrar todo lo que has mencionado, que en su mayoría ya se encuentra implementado.
    • It is unclear how to get from one sentence to the next and how to select them by language or other criteria...

      some sentence elements are considered as one single "word" also they aren't, but are a combination of words such as "it's" that should count as 2 words...

      Also, the final dot shouldn't be counted as a letter...
      • My plan right now is to embed this tool inside a page with navegation, filter and maybe even query tools.

        I'll try to see that elements such as dots, commas, etc. appear from the start in the sentence.
          • #
          • CK
          • CK
          • il y a 3 jour(s)
          I threw together a navigation system for using this script.

          1. Install the Script (http://userscripts.org/scripts/show/170945)
          2. Go to http://goo.gl/hH0lE
          3. There are over 11,000 French sentences, each with at least 1 English sentence.
          • Thanks, but the "next" button doesn't seem to work. I always see the same sentence when I click there, ie, "Une balle descend la rivière en flottant.".

            Is it possible to do it with the other languages?
        • Do you remember Braulio's Tatoeba Racing? Perhaps this could be done similarly:
          http://race.braulio.net.br/
      • the punctuation issue should be resolved if you reinstall the new version (same site)
    • As a core tool, it is very nice. The idea is great. With the intended improvements, I am sure it will be a highly enjoyable tool with a very user-friendly interface Thank you and keep up the hard work :)
      • Yes, a very good learning tool, indeed.
    • It's also an excellent tool to bust errors...
      • I was just about to say that!
  6. como puedo usar tatoeba
  7. I saw that this sentence:

    http://tatoeba.org/eng/sentences/show/33015

    had a "@needs native check" tag, so I looked at it. I agreed with the previous commenter that it was a sentence fragment, so I adopted it and made it into a full sentence. After that, I wanted to tag it "OK". But I got this message: "Tag 'OK' already exists for sentence #33015, or cannot be added".

    I find this message rather unhelpful. First of all, the code should be able to distinguish between those two cases. But more importantly, it would be good to explain WHY the tag can't be added. Even better would be if the tag COULD be added, since I don't see a reason why it shouldn't be allowed. Perhaps there's a policy that one can't mark one's own sentence OK? I would disagree with that policy, but if it is in place, there ought to be another way that one can signal that the requested native check was supplied.
    • > Perhaps there's a policy that one can't mark one's own sentence OK?
      Yes, there is. You could have added the tag before you adopted the sentence.
      • It seems a little problematic to me (especially because it makes following the logs confusing) that you need to mark the sentence "OK" before it actually is OK. But if this needs to be the case, I would argue that the message should say "You cannot mark your own sentence OK." This would be more informative.

        • It is standard policy that authors can't OK tag their own sentences. Otherwise, OK tags would lose their point...
          • Yes, I think that's evident. This tag doesn't mean "I (the owner of the sentence) think this sentence is OK" (We can take this for granted, can't we?). It means that at least a second person (and this one necessarily has to be a native speaker) confirms that the sentence is OK.
    • I basically agree with your point, but
      > I adopted it and made it into a full sentence.
      you shouldn't have done this, since the sentence had already Japanese and Turkish translations that are clearly not talking about Picasso.
      You can change an orphan sentence only when you're sure that it'll match all the translations or when you don't change the meaning, like when you only correct the punctuation.
      I'm not sure whether this is the shared policy, but at least this is what I believe.
      • Sorry. I wasn't set up to see Japanese and Turkish sentences, so I didn't realize they were there. I've changed my settings so that I now see all languages.

        I think it is okay to change the meaning of an orphan sentence (provided it's only a fragment or is otherwise problematic) as long as you unlink it from other sentences. Is that correct, administrators? I really hope it is, since many bad orphan sentences attract at least one translation, and some of these sentences are fragments or don't have a clear meaning to start with.
        • I personally prefer to leave a comment in the sentences linked to it (maybe the whole set of sentences are just fragments or have the same kind of problem as the orphan one) and not unlinking unless they have already some translation and I'm sure they don't match, in order to give the owner the chance to correct their sentence if they want to.
          • C'est une question extrêmement délicate, sans doute la plus délicate de Tatoeba.
            Question de contexte...
            On peut être persuadé que deux phrases ne correspondent pas, bien qu'étant "expert" dans les deux et, pourtant, avoir tort parce qu'on ne se place pas dans le bon contexte. Certaines phrases sont tellement vagues qu'elles peuvent être traduites dans des dizaines de sens différents alors qu'à priori, on n'en voit qu'un qui soit évident...erreur grossière !

            Je considère que délier des phrases subtilement liées, contre l'évidence, constitue le plus grand crime contre Tatoeba.
            • Moi, je parlais d'un contexte comme le suivant:
              On a une phrase anglaise incorrecte (on peut bien prendre la phrase dont Alan parlait "I think it's very like."), qui a été traduite au turque avec le signifié "je crois que c'est très similaire" mais alors la phrase en anglais change car elle n'était pas correcte et Alan n'a pas vu la phrase turque; après le changement fait qui a ajouté quelques mots parmi lesquels on trouve le nom "Picasso", il n'y a aucune possibilité que ces deux phrases soient similaires ou doivent être liées. C'est dans ce cas extrême où il n'y a aucune possibilité d'en être tort et où on peut délier (surtout si la phrase turque a été traduite).

              C'est bien pour les raisons dont tu parles que j'ai écrit dans mon commentaire "and I'm sure they don't match", dans mon cas, si je ne suis pas absolument sûre (et j'ai toujours constance de mon ignorance sur toutes les langues, la mienne incluse, alors il me faut des phrases comme celles dont je parlais dans le paragraphe précédent pour en être sûre), je ne vais pas délier. Si je crois que ça peut être le cas, je laisserai un commentaire sous la phrase qui me semble ne pas être une bonne traduction et attendrai une réponse avant délier.
              • Ben il ne faut pas traduire les phrases anglaises incorrectes...
        • I think the best would be

          1 - correct / complete
          2 - unlink
          3 - put a comment on the unlinked sentence, so that the other can correct and reask for linking (I think we have enough advanced contributors / moderators so that "ling again" shouldn't be a big deal)

          the advatange is that if the linked sentence owner does not anwer, we don't finish with sentences linked though they shouldn't be
          • That makes sense. I will think about writing this information somewhere on the wiki. This fits into a category of questions that aren't exactly FAQ (because they're not asked frequently), but come up often enough to be addressed.
  8. [fra] Rappel !
    Merci de ne pas traduire les phrases orphelines (qui n'ont pas de propriétaire) car elles sont, par nature, suspectes puisque personne ne veut en assurer la paternité. La plupart sont des phrases peu naturelles ou médiocres, provenant du Corpus Tanaka.

    [epo]
    Bonvolu ne traduki orfajn frazojn (tiuj kiuj ne havas posedanton) ĉar ili estas kompreneble suspektindaj kiel neniu volas preni la patrecon de ili.
    La plejmulto de ili estas ne-naturaj aŭ malbonaj, devenante de la Tanaka frazaro.

    [eng] Reminder !
    Thank you not to translate orphaned sentences (these that have no owner) because they are, by nature, suspect since nobody wants to assume responsibility for them. Most of them are not so natural or poor, sourced from the Tanaka Corpus.
      • #
      • CK
      • CK
      • il y a 28 jour(s)
      I've read through almost all the English sentences from the Tanaka Corpus, only ignoring the very, very long entries which are not useful for the kinds of projects I use the Tatoeba Corpus for.

      I adopted the ones that I felt were useful for my students and sounded natural, and added OK tags to ones that were already adopted by others.

      There are likely a few OK sentences left in this group, but the percentage is likely very low.

      There will be a higher percentage of good unadopted Japanese sentences, since they haven't yet all been read by our Japanese native speakers.

      Bunbuku is systematically going through them, but it's a time-consuming job and she seems to be the only one doing it at this time, so it will take a while.

      When it comes to Japanese, an unadopted sentence is perhaps just as likely be correct as one submitted by a non-native Japanese speaker.
        • #
        • CK
        • CK
        • il y a 5 jour(s)
        >When it comes to Japanese, an unadopted sentence is perhaps just as likely be correct as one submitted by a non-native Japanese speaker.

        Actually, perhaps I should have said "more likely to be correct than ones submitted by non-native Japanese speakers." The original Tanaka Corpus was created by Japanese students.
        • I think it's quite right to say that a Japanese sentence from the Tanaka Corpus that is left unadopted till now is as likely to be good as one submitted by a non-native Japanese speaker.
          It seems to me that many of the students did not try to write natural-sounding translations.
            • #
            • CK
            • CK
            • il y a 5 jour(s)
            >It seems to me that many of the students did not try to write natural-sounding translations.

            That's true, but they didn't make the obvious collocations errors that are being made by some of the current non-native members who like to contribute in Japanese. Japanese students also, for the most part, didn't make mistakes in the usages of を、に、が、は、etc.

            Quite often, the only unnaturalness is in the overuse of かれ、かのじょ、わたし、etc.

            • Unobvious errors are more harmful because learners are less likely to notice them.

              Moreover, there are quite a few sentences that have problems other than pronouns. I've even seen some sentences with unnatural てにをは. For example, they sometimes used "を" just because the English had a direct object, where actually "は" is more appropriate.

              Anyway, it's no use arguing with you here. I guess we both share the hope that Tatoeba will stop displaying all the orphan sentences to normal users.
                • #
                • CK
                • CK
                • il y a 5 jour(s)
                I think one point I was trying to make is that non-native speaker sentences are often as bad as or worse than orphan sentences.
    • [eng]

      I agree with CK's comment about orphan Japanese sentences.
      I want to make Tatoeba a good tool to study Japanese, so I'm eager to know what learners of Japanese think about all those unadopted sentences.

      [jpn]

      アドプトされていない日本語の文については、ぼくもCKさんのおっしゃる通りだと思います。
      ぼくはタトエバを日本語学習に役立つサイトにしたいと思っているので、日本語学習者のみなさんがこれらの文についてどういうふうにお考えなのか、ぜひとも知りたいです。
      • I come across unadopted Japanese sentences fairly often. For the most part, they don't seem too bad and I use them as if they were correct. (As far as translating them: I can't remember coming across an unowned Japanese sentence which didn't already have an English translation, so no worries there.) I have run across some iffy sounding sentences, and I usually post a comment on those asking if they are correct or not.
        • #
        • kuma
        • kuma
        • il y a 27 jour(s)
        It's good to hear that about unadopted Japanese sentences. I usually tend not to pay them any attention when looking for vocab examples, just in case they're wrong.
          • #
          • CK
          • CK
          • il y a 27 jour(s)
          kuma,

          This is still a safe policy. There are quite a few native-speaker owned Japanese sentences that you can use. Rather than learn from potentially poor sentences examples, it's probably safest to only study and translate from sentences owned by native speakers, and ignore unowned sentences and non-native produced sentences.
      • I've had good results using unowned Japanese sentences to study. The error rate that I've observed is quite low. On rare occasions I look at my flash cards later, realize something is strange, and ask a Japanese friend to see what they think. But this has only happened maybe 5 times in the last year, so overall I'm rather happy.

        (Also, the Japanese native speakers on this site who answer my questions and create example sentences have been a big help. I feel very lucky that they are around and active.)
      • I used to translate unadopted Japanese sentences I could understand. But then, after some alarming reports about the quality of those sentences Tommy and other Japanese members submitted, I decided to translate only the sentences created/owned by native speakers. That is, I avoid translating orphaned sentences and those owned by non-native speakers no matter how correct they look to me. As Tommy repeatedly pointed out in his messages, even grammatical Japanese sentences may still be absolutely unnatural and, thus, not just useless for a language learner (for a beginner, especially) but rather harmful. My Japanese level is simply not high enough to determine whether the sentence is good or not, so, as things now stand, I prefer to refrain from using/translating orphaned sentences and sentences owned by non-native speakers unless they tagged OK.
      • Ich finde viele der verwaisten japanischen Sätze seltsam und würde mir wünschen, daß Muttersprachler diese nach und nach durch natürliche, stilistisch einwandfreie Sätze ersetzten. Zum Beispiel hätte ich es für gut befunden, diesen Satz:

        http://tatoeba.org/deu/sentences/show/152508

        ohne viel Federlesens durch diesen:

        http://tatoeba.org/deu/sentences/show/2449219

        zu ersetzen, statt einen neuen hinzuzufügen.

        Meine grundlegende Meinung: Sätze, die zeigen, wie man sich möglichst nah am Englischen ausdrückt, sind nicht lehrreich und gehören getilgt.

        Trang hat mir in einer privaten Nachricht angekündigt, sich bald dazu zu äußern. Ich warte voller Ungeduld. ☺
      • Thank you all for your comments.

        I asked this because I thought there might be someone who's somehow making good use of those doubtful sentences, but it seems there aren't any. I'm still interested in whether there are any fans of orphan sentences (of any language). Please give me a reply if you're one of them.

        To those who don't distinguish between owned and unowned sentences, I strongly recommend you stop using Tatoeba that way unless you really want to waste your time. If you think that most of the orphan Japanese sentences are fine, that means you're not good enough at Japanese to judge. There are lots and lots of unnatural sentences that are literal translations of English, and many of them resemble the Japanese that beginners speak, so it might be especially difficult for native English speakers to find out what's wrong. It's true that most of them are good enough to make yourself understood. If you use them, most Japanese would understand you and say ニホンゴガオジョウズデスネー. But even so, serious learners must not try to learn from them.

        I'd recommend any users interested in this topic to read this thread where we discussed the similar matter three months ago:
        http://webcache.googleuserconte...clnk&gl=jp

        @Pfirsichbaeumchen
        I often don't feel like replacing an orphan sentence with my new one when I'm not sure if the sentence is really bad. You seem to trust us too much... we are only ordinary Japanese speakers. We can say "I wouldn't use this," but often we can't say "This is wrong." There might be lots of people who actually use the phrase I don't use. This is actually quite likely to happen since the divergence among speakers is significant. That's why I sometimes prefer leaving the original one as it is.

        I guess you want us to reduce the number of orphan sentences displayed on the site, but it'll take us centuries to correct the 123,741 sentences. I believe the only realistic way to protect learners from harmful sentences is, as I've written many times, to stop displaying them by default. If there is any serious reason that everybody should see them, then I'd suggest adding a warning message on every page of an orphan sentence such as "This sentence has no owner, which means its quality is doubtful. It could be harmful for your brain. It could even kill you!"
        • If the administrators support the idea that unowned sentences should not be translated, then I would think they should implement this policy via the user interface. Administrators, is there any reason why you wouldn't want the site software to prevent translation of an unowned sentence?
          • We've had some discussions similar to this one. I hope sysko can restore the wall so you can read them.
            • Would you be able to summarize the conclusions that you remember, the names of any of the participants, or when the discussions took place? Even if the wall were restored, it would be pretty hard to do a blind search for such a discussion. This is another reason why I'm really looking forward to the wiki going online.
              • > it would be pretty hard to do a blind search for such a discussion

                Not with a good Google search. :-)
                It's thread #15321.

                It's not exactly about orphaned sentences, but some of the ideas discussed there would be able to solve this issue, too.
              • It's here.
                http://webcache.googleuserconte...clnk&gl=en

                And "the other thread" sysko mentions is this one.
                http://webcache.googleuserconte...clnk&gl=en

                I'm really looking forward to the next version of Tatoeba, but since it seems that it takes a while until it completes, I think it'd be nice if something can be done for the orphan sentences with a little modification to the present version.
                • Thanks for the links!
        • When deciding whether to use an unowned sentence or not, the first question I ask is: is there a native-owned sentence that uses the same word I'm trying to learn? If so, I use that. If not, the unowned sentence will usually do.

          For me, it's about the words, and learning the words in context, even with questionable "beginner" grammar is more useful to me than learning the word without anything around it. Of course, natural grammar is the best and that's what I search for, but as a last resort, at least I can get a feel for the meaning of the word even if a native "wouldn't say it like that."
          • In that case, you can make a list of words and show it to us, then we can adopt or make some sentences using the words. We're here to help you. That's one of the greatest thing in Tatoeba, isn't it? ^^
            We won't be able to respond to all the requests, to be sure, but we'll do what we can.
            I wish there were a useful system to do this kind of thing easily.
            • That's a good idea. When we have a suitable place (perhaps on the new wiki) for tracking requests, we should add it there.
        • I like the idea of orphan sentences not showing up under searches. (Perhaps a settings option or a search query check box or something like that would be suitable.)
          • I, too, have come to somewhat like that idea. Another possibility would be to show an easy-to-spot warning sign next to an orphan sentence, especially when it is displayed as a (direct or indirect) translation.
  9. Another suggestion (because I'm in the process of listing ones that have occurred to me before): it would be nice to have an function that lets you search for existing duplicates or near-duplicates of a sentence that you're about to submit, ideally without disturbing your workflow. (Thus, displaying this information in a separate frame would be helpful.)
    • Automaticamente, no se como se podria hacer. Pero he pensado que seria util un sistema para relacionar frases, diferente del de las etiquetas. Seria como etiquetas o grupos sin nombre.

      Por ejemplo, "a Tom le han despedido" se podria relacionar con "A tom le han contratado"

      Se pueden hacer etiquetas, pero a veces no sabes que nombre darle o son unas pocas frases, que no merece ponerles una etiqueta, pero es util que te sugieran cuando ves alguna de las frases relacionadas.

      Esto con ejemplos se veria mucho mejor, a ver si busco unos ejemplos.

      Con la version actual de tatoeba, no se puede pero en la siguiente se pueden utilizar el sistema de enlaces entre frases para mas cosas que unir tradiciones entre idiomas.
    • Espero que se me entienda. Tal vez he sido un poco lioso.
      • I did understand what you said. It makes me realize that my original comment included several requests:

        (1) Provide a search that automatically looks for fuzzy matches. The current search engine (Sphinx) doesn't do this.

        (2) Provide functionality that allows you to perform a search through the corpus for the content that you're about to submit. If part 1 is implemented, then this search can find similar as well as identical sentences.

        Currently, in order to search for content that you're about to submit, you basically need to keep a separate open Tatoeba window, then copy your content (wrapped in double quotes if you want to find an exact match) into the Tatoeba search bar. This takes up extra screen space and of course also requires extra manual effort. As a result, I suspect that few people look for duplicate or related sentences on every submission. (Realistically speaking, it probably only makes sense to do these searches on short sentences with common vocabulary.) There is a script that is run periodically on the corpus to merge exact duplicates, so adding them is not as problematic as it might be. However, it's still a pretty complicated task to merge comments and histories as well as content, so avoiding the insertion of duplicates in the first place is preferable, especially if you're an advanced contributor, which gives you the ability to link to existing sentences.

        It sounds like you're talking about an ability to manually create a new kind of metadata link that would aid fuzzy searches by including searches for synonyms, rather than relying on some existing synonym dictionary to do the fuzzy search. Is that right?
  10. I recommend that we put links to "Help" and "FAQ" at the top of the page, where they'll be much more visible than at the bottom. (I didn't know they existed there until someone pointed them out to me.) I would vote that we also put the other items (such as "Terms of use" and "Contact us") at the top of the page, but it's particularly important to make "Help" and "FAQ" visible.
    • > I recommend that we put links to "Help" and "FAQ" at the top of the page, where they'll be much more visible than at the bottom.

      That's really very important. Believe me, sometimes people are contributing during a long period of time, before they notice this bare at the bottom. You have to scroll a long way down to get there (working for tatoeba usually you seldome have a reason to do so), and even than the bar there is far from eyecatching.
    • I think top right is good place. And maybe in the top of wall is usefull a explansion for begginer contributors and tricks.

      When you become a advance contribuitor, may usefull a mail with new knoledge like, link sentences by modifing numbers in the link.
<<
<
12345 >>>