Hi.
Speaking of Wikisource software, do we already have any instance set up for
OCR as a service? (I'm thinking of OpenOCR[1] hosted in a Docker[2]
container somewhere, perhaps on Labs.)
If yes, where is it and who maintains it, and can I use it as a client?
If not, I am prepared to set this up.
A.
[1] http://www.openocr.net/
[2] https://www.docker.com/
--
Asaf Bartov
Wikimedia Foundation <http://www.wikimediafoundation.org>
Imagine a world in which every single human being can freely share in the
sum of all knowledge. Help us make it a reality!
https://donate.wikimedia.org
Just to inspire true developers: I'm playing with the code of a new tool,
that replaces/adds diacritical marks of the character preceding the cursor
into edit textarea. The trick is to decompose the character with
.normalize("NFD"), to change what needs to be changed, and to recompose it
with .normalize("NFC").
Much faster and comfortable than the usual, painful search for the needed
exotic character into Special characters or into a Unicode table.
The code is primitive and undocumented, styling is horrible, but it runs
and it seems to be a stand-alone code running into any project ( NOT into
Visual Editor, I presume) , to test it simply mw.loader.load() it in edit
mode from
https://it.wikisource.org/wiki/Utente:Alex_brollo/Unicode_combining_diacrit…
and feel free to use/change/ignore it as you like: it's simply a "running
idea".
Alex
I agree with Billinghurst that user-made translations do not necessarily have to be dealt with the same way published editions/translations are. However, if the user-made translation is connected to the same WD item as the original work (edition, actually), then we should indicate which of the Wikisource links really leads to the text the statements refer to (i.e., the original work). This can easily be done with property 953, "full text available at".
Erasmo Barresi
> Date: Tue, 28 Jul 2015 07:59:46 +0000
> From: billinghurst
> To: "discussion list for Wikisource, the free library"
> , novartza(a)gmail.com
> Subject: Re: [Wikisource-l] more to edition(s) <-> edition or
> translation of
> Message-ID:
>
> Content-Type: text/plain; charset="utf-8"
>
> Nahum,
>
> I suppose the predominant part of my discussion was focusing on published
> translations, rather than WS-generated translations.
>
> 1) because many famous works have (famous) published translations, each
> with their own translator(s), date of publication, publisher, ... all
> worthy of noting and WD'ing.
>
> 2) you can only link to one work per wiki at WD, and as an example of an
> issue , I know of numbers of Chekov's works that enWS hosts where we have
> multiple published translations. So enWS needs a means for link generation
> for many to one (at ruWS), or for English language works, we may need a one
> to any.
>
> When it comes to WS-generated translations, I can see that we may or may
> not wish to call those translations their own edition. If a translation
> passes a notability check of its own edition, then separate, if not, maybe
> then it comes off the other original language work. (Maybe leave that
> decision with the xxWS to determine???).
>
> What I am hoping to see is a tool that provides the maximum flexibity and
> credibility for the WSes, yet gives best visibility and authority at WD.
>
> Billinghurst
>
> On Tue, 28 Jul 2015 12:35 Nahum Wengrov wrote:
>
>> Wait a minute.
>> If a work esists in, say, ru.wikisource,
>> And then someone translates that work and posts his translation under a
>> free license in he.wikisource,
>> I am not to link the hebrew version to its source in ru.ws on wikidata,
>> But to create a seperate wikidata entry for it?
>> This makes zero sense to me, and we never did it this way on he.wikisource.
>>
>> On Sun, Jul 26, 2015 at 2:25 PM, billinghurst
>> wrote:
>>
>>>
>>> Thought that the pasted discussion from WD is of interest and adds to our
>>> recent discussion on interwikis/interlanguage links.
>>>
>>>
>>> https://www.wikidata.org/wiki/Wikidata:Project_chat#Wikidata_discourages_in…
>>>
>>> Wikidata discourages interwiki links
>>>
>>> Looking at Zhuangzi (Q1074987)
>>> *Zhuanhgzi*, I see that this data item is only for the work as it exists
>>> in Chinese. The work as translated into English is a separate data item,
>>> and the French Wikisource translation is another data item.
>>>
>>> Effectively, this means that Wikidata discourages interwiki links to and
>>> between Wikisource projects, because they will never be part of the same
>>> data item. Further, anyone seeking a translation of a work into another
>>> language must first come to Wikidata and surf the links even to find out if
>>> translations of a work in another language exist on Wikisource; it is not
>>> possible to do that from any Wikisource directly.
>>>
>>> I thought the whole point of moving the links to Wikidata was to promote
>>> connections between projects, not to eliminate them. But perhaps I am
>>> wrong. --EncycloPetey (
>>> talk ) 02:19, 26
>>> July 2015 (UTC)
>>> @EncycloPetey : There
>>> are other ways to create interwiki by the help of Wikidata. See s:sv:Bibeln
>>> 1917 where I have made
>>> some tests with the Bible (Q1845) .
>>> The interwiki is created by the help of a Lua-module that follows edition(s)
>>> (P747) and edition or
>>> translation of (P629) . The
>>> big advantage is that it makes possible to create intewiki to more than one
>>> page in every project. For example, that page have 13 links to
>>> enwikisource. -- Innocent bystander
>>> (talk
>>> ) 05:49, 26
>>> July 2015 (UTC) @Innocent bystander
>>> :This is very
>>> interesting, what template (and Lua-module) do you use ? it should be done
>>> for all wikisources, that often have a lot of translated texts :) --
>>> Hsarrazin (talk
>>> ) 09:43, 26 July 2015
>>> (UTC) @Hsarrazin : It's s:sv:Modul:Sandlåda/Innocent
>>> bystander
>>>
>>> who today is included in s:sv:Mall:Titel
>>> , a template that can be
>>> found in almost every page on svsource. Observe that the module is not
>>> secured against loops in the P747/P629-hierarcy. It also needs support by
>>> the "interwiki-extra"-class in s:sv:MediaWiki:Common.js
>>> , otherwise you
>>> cannot have more than one interwiki in each page. -- Innocent bystander
>>> (talk
>>> ) 10:11, 26
>>> July 2015 (UTC)BTW, another thing the code does: It makes it possible to
>>> have interwiki to the Text-namespaces in als/bar/frr/pflwikis. -- Innocent
>>> bystander (talk
>>> ) 10:18, 26
>>> July 2015 (UTC)
>>>
>>> _______________________________________________
>>> Wikisource-l mailing list
>>> Wikisource-l(a)lists.wikimedia.org
>>> https://lists.wikimedia.org/mailman/listinfo/wikisource-l
>>>
>>>
>> _______________________________________________
>> Wikisource-l mailing list
>> Wikisource-l(a)lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/wikisource-l
>>
The problem with item Q936276 is that three of the four "editions" are not editions at all--one is a TV series and two are films. I just asked on Wikidata (https://www.wikidata.org/wiki/Property_talk:P144#Reverse_property.3F) what property should be used in such cases.
Erasmo Barresi
> Date: Thu, 16 Jul 2015 01:01:53 +0000
> From: billinghurst <billinghurstwiki(a)gmail.com>
> To: "discussion list for Wikisource, the free library"
> <wikisource-l(a)lists.wikimedia.org>
> Subject: [Wikisource-l] More on difficulties of discovery of WS within
> Wikidata
> Message-ID:
> <CAK97GTdwHF6eZwwjx3FaiPn0T5P=6KuxG-t9Abead8T3R3rH-w(a)mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> I was just working through data bits and pieces and found another
> interesting example of the issues of difficulties of separation.
>
> d:q936276 is the book "wind in the willows" from there find out which one
> has the edition of the book.
>
> While I know that wikidata is primarily focused on data pulls, that there
> is no visual cues is an issue. One wonders whether there is an opportunity
> to have sister flags against each of the "edition(s)".
>
> Regards, Billinghurst
The Second International Conference on Computer Graphics, Multimedia and
Image Processing (CGMIP2015)
Universitas Siswa Bangsa Internasional
Jakarta, Indonesia
October 29-31, 2015
http://sdiwc.net/conferences/cgmip2015/
cgmip15(a)sdiwc.net
----------------------------------------------
CALL FOR PAPERS, WORKSHOP and SPECIAL SESSIONS
----------------------------------------------
Research topics not limited to:
- 3D imaging
- Case studies and emerging technologies
- Color and texture
- Computational geometry
- Computer art and entertainment (including games)
- Case studies
- Classification and clustering techniques
- Compression methods
- Computer animation
- Curves and meshes
- e-Learning applications and computer graphics
- Emerging technologies
- Emerging display technologies
- Face recognition, face detection, and gesture recognition
- Fractal geometry and applications
- Graphics algorithms and applications
- Graph theory in image processing and vision
- Human-computer interfaces
- Image compression, coding, and encryption
- Image feature extraction
- Image geometry and multi-view geometry
- Immersive virtual reality
- Interactive digital media
- Simulation and virtual reality
- Software tools for virtual reality
- Surface modeling
- Virtual environments
- Virtual laboratories
- Virtual reality, visualization, and education
- Visualization
All registered papers will be included in SDIWC Digital Library, and in
the proceedings of the conference. All submitted papers will be reviewed
by a minimum of two reviewers. The published proceedings will be indexed
in ResearchBib, in addition, they will be reviewed for POSSIBLE
inclusion within the INSPEC, EI, DBLP, Microsoft Academic Research,
ResearchGate, and Google Scholar Databases.
Important Dates
===============
Submission Deadline: September 29, 2015
Notification of Acceptance: 2-4 weeks from the submission date
Camera Ready Submission: October 19, 2015
Registration Deadline: October 19, 2015
Conference Dates: October 29-31, 2015
I was just working through data bits and pieces and found another
interesting example of the issues of difficulties of separation.
d:q936276 is the book "wind in the willows" from there find out which one
has the edition of the book.
While I know that wikidata is primarily focused on data pulls, that there
is no visual cues is an issue. One wonders whether there is an opportunity
to have sister flags against each of the "edition(s)".
Regards, Billinghurst
For some reason when I uploaded the zip file with images to the internet
archive, the djvu was not generated... does anyone know the reason?
The zip file can be found here
https://archive.org/details/ElsMestresContistesVolum1
Cheers,
Micru