Just to note, just for the time being, there is even a blessing hidden in disguise for the lackof a proper OCR for Indic languages:
Although it may look ridiculous at first, the very reason that we do not have an OCR forced us to the conduct the 2014 Digitization contest in ml.wikisource. More than milking thousands of valuable and rare old pages, it served another purpose that was not originally intended for: thousands of school children as well as otherwise non-wikimedian volunteers learned and enhanced their typing skills in local languages!
OCR is a great and urgent necessity for all Indian languages. But that is not always an excuse too. We can still pursue our mission even through such constraints, leveraging every disadvantage into an advantage.
When you have a great amount of work such as 8376 pages, do not be overwhelmed! Think of the Moore's law. And in India, our strength is our population. We will never run out of work. But we will possibly also never run out of volunteers if sufficient environment is prepared and maintained for them.
The community power of India's digital future is yet to arrive. But it is just around the corner, for sure. We just need to set ourselves ready for that. So, keep thinking of possibilities and shake off limitations!
:)
-Viswam
On Sun, Aug 31, 2014 at 3:38 AM, Jayanta Nath jayantanth@gmail.com wrote:
What is the present status of Tamil OCR? 20 volumes (8376 pages) typing should not solution for this!
Jayanta
On Sun, Aug 31, 2014 at 1:07 AM, ViswaPrabha (വിശ്വപ്രഭ) < viswaprabha@gmail.com> wrote:
Let me mention of yet another possible model that we are almost about to try in ml:
Apart from the already tried and tremendously successful models of school-wise student model and general community competition (for prizes in kind or just for credential certificates), we are now pondering upon a new idea:
Use secondary social network collaboration.
As an example, if you have formed (or if you could grow from now,) a very vibrant Wikimedia or other language-focused community network in Facebook, you can open up an appropriately scaled competition model within that. Many people are now seeking independent localization tools and input methods just to use Facebook.
(And BTW, thanks to Facebook, which I always detested, for making a great invisible revolution among mainstream local community members, that most other models are still struggling to achieve!)
In Malayalam, now we have several such large communities each one specializing in their own arenas (eg. Butterflies, Plants, Birds, Grammar, Films and Film Songs etc. etc.). They all share the idea that most of such shared knowledge and digital text and images shall ultimately end up in WM pools.
So, imagine a world where, A Tamil community in Facebook takes up this project. (May or may not be with an organized competition spirit). Each member picks up a few pages one at a time, and post it back to Facebook. A small team collects this and pipes to Wikisource. Eventually, the volume gets completely in.
Just imagine... and it shall happen!
PS. Why in FB and then why not directly in Wikisourse? You know why! The interface makes a big difference. Let's admit that!
-Viswam
On Sun, Aug 31, 2014 at 12:03 AM, Yann Forget yannfo@gmail.com wrote:
FYI. Yann
---------- Forwarded message ---------- From: Ravishankar ravidreams@gmail.com Date: 2014-08-30 20:10 GMT+05:30 Subject: [Wikimediaindia-l] 20 volumes (8376 pages) of Tamil Encylopedia released under Creative Commons To: Wikimedia India Community list <wikimediaindia-l@lists.wikimedia.org
Hi,
Tamil Development Board (an autonomous institution under Government of Tamilnadu) releases its Encyclopedia (10 volumes, 7407 pages) and Children's Encyclopedia (10 volumes, 969 pages) under Creative Commons license. Tamil Wikipedians lead by Prof. C. R. Selvakumar and Prof. P. R. Nakkeeran, (Director, Tamil Virtual Academy) spearheaded this initiative coinciding with Tamil Wikipedia's 10 years celebrations.
An official confirmation (in Tamil) can be seen at
https://upload.wikimedia.org/wikipedia/commons/4/46/Letter_from_Tamil_Develo...
Scanned copies of these works are already available at
http://tamilvu.org/library/kulandaikal/lku00/html/lku00ind.htm
At Tamil Wikipedia, we are discussing how we can get this content typed and transferred to WikiSource. Doing so can be a good model to encourage more such works to be released in public domain.
Following are two options I can think of:
- Volunteers type all the content. Besides taking years to complete,
this won't do justice for the value of time of volunteers who can do more valuable work than typing mechanically.
A program like IT@School present in Kerala or a contest can encourage more people to join this effort but not all communities can't emulate this model successfully.
- Request WMF to give a grant to the owner of the content and let
them hand over the typed content to Wikisource volunteers who will upload and wikify the content.
This will ensure maintaining the spirit of volunteerism and yet getting the work done in a professional and time bound manner.
Numerous works in Wikisource are such ready made content uploaded already in the web through other projects like Project Gutenberg.
If providing grants to non-Wikimedia organizations is an issue, a grant towards this can be given to community / chapter who will then outsource the typing work.
I welcome community's input on any other model for this as India has vast amount of literature and works like this are waiting to be transfered to Wikisource. This is one area where we can add lot of content to Wiki projects at once.
Ravi
Wikimediaindia-l mailing list Wikimediaindia-l@lists.wikimedia.org To unsubscribe from the list / change mailing preferences visit https://lists.wikimedia.org/mailman/listinfo/wikimediaindia-l
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
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l