If you look at this chapter of a novel in the French language, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1
...it has a link to the corresponding chapter in Russian and next to the word "Русский" is a little arrow ⇔ or <=>
If you click that arrow, you get the two texts placed side-by-side, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1?match=ru
But there is no further matching. Paragraphs are not next to each other. If I understand correctly, this is the "Double Wiki" extension, https://www.mediawiki.org/wiki/Extension:DoubleWiki https://wikisource.org/wiki/Wikisource:DoubleWiki_Extension
Is anybody using this feature in a serious way? Does it have any more details that can make the matching better? If both texts had numbered paragraphs and sentences (something like the Bible), it would in theory be possible to match them by number.
very few improvements since it was created. :( https://phabricator.wikimedia.org/diffusion/EDWI/history/master/
On Mon, 22 Aug 2016 11:34 Lars Aronsson, lars@aronsson.se wrote:
If you look at this chapter of a novel in the French language, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1
...it has a link to the corresponding chapter in Russian and next to the word "Русский" is a little arrow ⇔ or <=>
If you click that arrow, you get the two texts placed side-by-side, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1?match=ru
But there is no further matching. Paragraphs are not next to each other. If I understand correctly, this is the "Double Wiki" extension, https://www.mediawiki.org/wiki/Extension:DoubleWiki https://wikisource.org/wiki/Wikisource:DoubleWiki_Extension
Is anybody using this feature in a serious way? Does it have any more details that can make the matching better? If both texts had numbered paragraphs and sentences (something like the Bible), it would in theory be possible to match them by number.
-- Lars Aronsson (lars@aronsson.se) Linköping
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
I think the idea is brilliant, but DoubleWiki is really old and the arrow is almost invisible, so probably no one is using it.
When ContentTranslation came out I wondered for a while if that could be applied also to Wikisource (for translations) or if part of it could be used to match existing translations we already have on Wikisources...
Aubrey
On Mon, Aug 22, 2016 at 7:01 AM, jayvdb@gmail.com wrote:
very few improvements since it was created. :( https://phabricator.wikimedia.org/diffusion/EDWI/history/master/
On Mon, 22 Aug 2016 11:34 Lars Aronsson, lars@aronsson.se wrote:
If you look at this chapter of a novel in the French language, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1
...it has a link to the corresponding chapter in Russian and next to the word "Русский" is a little arrow ⇔ or <=>
If you click that arrow, you get the two texts placed side-by-side, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1?match=ru
But there is no further matching. Paragraphs are not next to each other. If I understand correctly, this is the "Double Wiki" extension, https://www.mediawiki.org/wiki/Extension:DoubleWiki https://wikisource.org/wiki/Wikisource:DoubleWiki_Extension
Is anybody using this feature in a serious way? Does it have any more details that can make the matching better? If both texts had numbered paragraphs and sentences (something like the Bible), it would in theory be possible to match them by number.
-- Lars Aronsson (lars@aronsson.se) Linköping
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
On 08/22/2016 09:18 AM, Andrea Zanni wrote:
I think the idea is brilliant, but DoubleWiki is really old and the arrow is almost invisible, so probably no one is using it.
Those who could use such a function are parallel corpus linguists. "Opus" is a project that collects open-source parallel corpora, http://opus.lingfil.uu.se/ In the Czech Republic, the project InterCorp also does similar work, and they developed their own "InterText" editor software for the alignment, a paragraph and a sentence at a time, http://wanthalf.saga.cz/intertext The screenshot gives you an idea of their user interface, http://wanthalf.saga.cz/pictures/InterText_screenshot.png
But that user interface could be "DoubleWiki" if you incorporate some ideas from Mediawiki's visual editor...
Now, do we have any corpus linguists who use Wikisource? I think we need to have the input from the actual users.
On Sun, Aug 21, 2016 at 9:34 PM Lars Aronsson lars@aronsson.se wrote:
If you look at this chapter of a novel in the French language, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1
...it has a link to the corresponding chapter in Russian and next to the word "Русский" is a little arrow ⇔ or <=>
If you click that arrow, you get the two texts placed side-by-side, https://fr.wikisource.org/wiki/Le_Voyageur_enchant%C3%A9/Chapitre_1?match=ru
But there is no further matching. Paragraphs are not next to each other. If I understand correctly, this is the "Double Wiki" extension, https://www.mediawiki.org/wiki/Extension:DoubleWiki https://wikisource.org/wiki/Wikisource:DoubleWiki_Extension
Is anybody using this feature in a serious way? Does it have any more details that can make the matching better? If both texts had numbered paragraphs and sentences (something like the Bible), it would in theory be possible to match them by number.
In most cases one sentence in the original does not correspond to one sentence in a translation. In many cases, one paragraph in the original does not correspond to one paragraph in a translation. As long as the Double_Wiki tool is a general-purpose tool for comparing different language versions, it's going to be massively limited in what it can do in these fashions.
wikisource-l@lists.wikimedia.org