I am cataloguing the non-WM contacts I made at Wikimania. Reading up on them, basically making the time I didn't have during the conference to understand who they are. The BHL had three librarians at the conference, I remember that they were very interested in a solution for getting OCR corrections back into djvu files. They mentioned Wikisource in the blog post they made about the conference. I thought the list might be interested in what they had to say.
http://blog.biodiversitylibrary.org/2012/07/wikimania-2012.html
Birgitte SB
Hey everybody,
On 04-Aug-2012, at 5:41 PM, Birgitte_sb@yahoo.com wrote:
I am cataloguing the non-WM contacts I made at Wikimania. Reading up on them, basically making the time I didn't have during the conference to understand who they are. The BHL had three librarians at the conference, I remember that they were very interested in a solution for getting OCR corrections back into djvu files. They mentioned Wikisource in the blog post they made about the conference. I thought the list might be interested in what they had to say.
http://blog.biodiversitylibrary.org/2012/07/wikimania-2012.html
I'm currently working with the BHL on a two-month, unrelated metadata project, part of which is making sure that BHL's illustration metadata can be easily synced with content in the Commons (see http://commons.wikimedia.org/wiki/Template:Information_Art_of_Life for more details). So I don't know anything about BHL's plans for Wikisource, but let me know if I can help!
Personally, I'd love to see more (annotated!) biodiversity texts in Wikisource, such as http://en.wikisource.org/wiki/The_Salticidae_(Spiders)_of_Panama/Zygoballus -- these species descriptions are the formal definition of a new species, and the BHL has been a huge help in making these definitions available, online and for free, to taxonomists everywhere (to say nothing of tons of gorgeous illustrations! [1]). However, their transcription and indexing are largely automated, via OCR and text matching. Moving these essential resources into Wikisource, where transcriptions and indexing could be improved by hand, would be awesome!
cheers, Gaurav http://en.wikisource.org/wiki/User:Gaurav
[1] http://commons.wikimedia.org/wiki/Commons:Biodiversity_Heritage_Library
On Aug 5, 2012, at 7:38 PM, Gaurav Vaidya gaurav@ggvaidya.com wrote:
Hey everybody,
On 04-Aug-2012, at 5:41 PM, Birgitte_sb@yahoo.com wrote:
I am cataloguing the non-WM contacts I made at Wikimania. Reading up on them, basically making the time I didn't have during the conference to understand who they are. The BHL had three librarians at the conference, I remember that they were very interested in a solution for getting OCR corrections back into djvu files. They mentioned Wikisource in the blog post they made about the conference. I thought the list might be interested in what they had to say.
http://blog.biodiversitylibrary.org/2012/07/wikimania-2012.html
I'm currently working with the BHL on a two-month, unrelated metadata project, part of which is making sure that BHL's illustration metadata can be easily synced with content in the Commons (see http://commons.wikimedia.org/wiki/Template:Information_Art_of_Life for more details). So I don't know anything about BHL's plans for Wikisource, but let me know if I can help!
Personally, I'd love to see more (annotated!) biodiversity texts in Wikisource, such as http://en.wikisource.org/wiki/The_Salticidae_(Spiders)_of_Panama/Zygoballus -- these species descriptions are the formal definition of a new species, and the BHL has been a huge help in making these definitions available, online and for free, to taxonomists everywhere (to say nothing of tons of gorgeous illustrations! [1]). However, their transcription and indexing are largely automated, via OCR and text matching. Moving these essential resources into Wikisource, where transcriptions and indexing could be improved by hand, would be awesome!
To be honest I don't think they knew what Wikisource did before Wikimania. So I doubt there is anything so firm as "plans". I know Aubrey and I, at least, spoke with them. They are definately interested in the platform at Wikisource, but they want to re-integrate the corrections made back into their collection. This is something that is a problem with djvu files that we do not yet have an answer for.
Birgitte SB
I'm currently working with the BHL on a two-month, unrelated metadata
project, part of which is making sure that BHL's illustration metadata can be easily synced with content in the Commons (see http://commons.wikimedia.org/wiki/Template:Information_Art_of_Life for more details). So I don't know anything about BHL's plans for Wikisource, but let me know if I can help!
Personally, I'd love to see more (annotated!) biodiversity texts in
Wikisource, such as http://en.wikisource.org/wiki/The_Salticidae_(Spiders)_of_Panama/Zygoballus-... these species descriptions are the formal definition of a new species, and the BHL has been a huge help in making these definitions available, online and for free, to taxonomists everywhere (to say nothing of tons of gorgeous illustrations! [1]). However, their transcription and indexing are largely automated, via OCR and text matching. Moving these essential resources into Wikisource, where transcriptions and indexing could be improved by hand, would be awesome!
To be honest I don't think they knew what Wikisource did before Wikimania. So I doubt there is anything so firm as "plans". I know Aubrey and I, at least, spoke with them. They are definately interested in the platform at Wikisource, but they want to re-integrate the corrections made back into their collection. This is something that is a problem with djvu files that we do not yet have an answer for.
Hi Gaurav, I agree with Birgitte :-) The BHL were very intrested in what we do, as were the guys from NARA. The GLAM is a huge and crucial "dimension" ofr Wikisource future, I tried (very, very badly) to list some things herehttp://meta.wikimedia.org/wiki/Wikimania_2012_Wikisource_roadmap#GLAM .
As Birgitte says, many GLAMs "want to re-integrate the corrections made back into their collection": Wikisource is a free human proofreading platform, we correct OCR which is still not perfect and won't be for many years. We "curate" texts and this is our added value.
So, one of the things Iyou could help us with is ask BHL *how* they would like to have the corrections reintegrated in their collection. Would they like the proofread text put back in the djvu? They want it *mapped*? it is crucial or they could be satisfied also with a simple text layer inside the djvu? Do they use djvu? Do they use other formats? Do they want something completely different?
These are some techical issues (which in part I'm discussing with Alex Brollo, who is digging again into the "put back the proofread layer innto the djvu" issue.
Another BIG issue is that we have a small userbase, and I someone comes to us and says "here, you can have a gazillion books", we won't proofread them untile the end of the times.
I think this is something who went "wrong" with the Gallica partnership (by far the greatest Wikisource-GLAM collaboration)(I'm not blaming anyone, the fr.source community did great, but the books were just too many). Is this something we can work together on? Reaching a critical mass of users is crucial for sister projects, and if some specific community found some solutions it would be great to share.
I'll start. One thing it worked fot it.source was setting up a "Proofread text of the month", which is also advertised in the Wikipedia Main page ( http://it.wikipedia.org/wiki/Pagina_principale). This lead some wikipedians to come and proofread few pages: we never had "casual" proofreaders and now we have :-)
Aubrey
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
On Aug 6, 2012, at 3:31 AM, Andrea Zanni zanni.andrea84@gmail.com wrote:
I'm currently working with the BHL on a two-month, unrelated metadata project, part of which is making sure that BHL's illustration metadata can be easily synced with content in the Commons (see http://commons.wikimedia.org/wiki/Template:Information_Art_of_Life for more details). So I don't know anything about BHL's plans for Wikisource, but let me know if I can help!
Personally, I'd love to see more (annotated!) biodiversity texts in Wikisource, such as http://en.wikisource.org/wiki/The_Salticidae_(Spiders)_of_Panama/Zygoballus -- these species descriptions are the formal definition of a new species, and the BHL has been a huge help in making these definitions available, online and for free, to taxonomists everywhere (to say nothing of tons of gorgeous illustrations! [1]). However, their transcription and indexing are largely automated, via OCR and text matching. Moving these essential resources into Wikisource, where transcriptions and indexing could be improved by hand, would be awesome!
To be honest I don't think they knew what Wikisource did before Wikimania. So I doubt there is anything so firm as "plans". I know Aubrey and I, at least, spoke with them. They are definately interested in the platform at Wikisource, but they want to re-integrate the corrections made back into their collection. This is something that is a problem with djvu files that we do not yet have an answer for.
Hi Gaurav, I agree with Birgitte :-) The BHL were very intrested in what we do, as were the guys from NARA. The GLAM is a huge and crucial "dimension" ofr Wikisource future, I tried (very, very badly) to list some things here.
<snip>
Another BIG issue is that we have a small userbase, and I someone comes to us and says "here, you can have a gazillion books", we won't proofread them untile the end of the times.
I think this is something who went "wrong" with the Gallica partnership (by far the greatest Wikisource-GLAM collaboration)(I'm not blaming anyone, the fr.source community did great, but the books were just too many). Is this something we can work together on? Reaching a critical mass of users is crucial for sister projects, and if some specific community found some solutions it would be great to share.
I think this is a very important issue. I don't know any tried and true
When I spoke the BHL troika, I particularly described Wikisource as a platform. I also explained to them that if they had any issues with working on our platform, they could optionally install MediaWiki and Proofread page on their own server as it was all open source. I did not wish to describe Wikisource as a *service* for proofreading and transcription. We are really not capable of being such a service. We *are* capable of providing the basic platform for transcription, of providing some guidance for presenting texts with very limited tech support, and also of handling the interwebs aspect (not only managing the servers and bandwidth but handling the user accounts, the privacy issues, the spambots, the drawing of lines that must be drawn somewhere, and the finding of space for the people who will inevitably show up at their project only to discover they would actually rather be doing something else entirely with it all).
I think sometimes we downplay too much what we are capable of offering, because we would wish to be able to offer more. Don't underestimate how valuable every piece of what we offer is. Just WMF and our community set processes taking responsibility for the privacy issues (and doing it competently by Internet standards) is an immense burden off of an institution! Many of these things are not issues anyone would necessarily know to think about before starting a crowd-sourcing project. But we should remember that they were (and still are) hard issues for the WM movement to figure out. We now have won some of the prize of age, the ability to draw on our own experience, and that of others within the movement, in order to spare those that partner with us from some of the problems of naïveté. This is meaningful. Even though we can never proofread all their documents with the efforts of just our small community, what we can do for them is meaningful.
NARA with their dashboard, directing people from their community straight into the Page: namespace of documents they want transcribed and proofread, is to my mind the way forward. I would like see this dashboard concept developed further, so we can show other interested institutions how to set dashboards up for their communities. So they might direct people they are connected with toward working directly with institutional texts on Wikisource without anyone getting too lost in in the wiki. I believe that we can be a platform for this sort work by these institutions, that we can offer institutional supporters a low barrier way to contribute directly to the mission they support, that we can help facilitate the institutional "followers" becoming a real community that does real work within the areas where our missions overlap. I don't believe we can, in good faith, accept a large donation of institutional digital files from institutions which expect the donation to be the beginning and end of their involvement with Wikisource.
Birgitte SB
wikisource-l@lists.wikimedia.org