Transcription annotations
208 views
Skip to first unread message
Antoine Isaac
Jul 20, 2017, 1:58:25 PM7/20/17
to IIIF Discuss, hugo.ma...@europeana.eu, sergiu...@ait.ac.at
Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
Thanks for the help!
Antoine
Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
Thanks for the help!
Antoine
Robert Sanderson
Jul 20, 2017, 4:41:28 PM7/20/17
to iiif-d...@googlegroups.com, hugo.ma...@europeana.eu, sergiu...@ait.ac.at
The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.
There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
HTH
Rob
Antoine
--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
--- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Rob Sanderson
Semantic Architect
The Getty Trust
Los Angeles, CA 90049
Message has been deleted
KANZAKI Masahide
Jul 21, 2017, 7:27:20 AM7/21/17
to IIIF Discuss
Hello Antoine, all
I have an experiment that connects IIIF images and TEI/XML transcription in Bodleian First Folio [1].
It does not embed link information in the manifest file. Rather, it uses separate Web Annotation description where the targets are images and bodies are TEI pages [2]. Although the targets are not canvases, it could be easily translated into IIIF AnnotationList.
There was a discussion in TEI mailing list on how to relate TEI and corresponding facsimile (image) [3].
hope this would be relevant.
cheers,
2017年7月21日金曜日 18時13分17秒 UTC+9 Antoine Isaac:
Hi Rob, all,
Thanks for the input!
By "re-use" I mean that the statements that represent transcriptions could be used in different contexts (IIIF, Text modeling) with as few changes as possible. Especially, while the target may be specific (the IIIF context expects IIIF resources like canvases) the other components of the annotation could remain the same. Especially the body, and - I was hoping - the motivation.
I reckon it may be tricky to blend a strict 'presentation' perspective with an approach that goes more into modeling the transcription 'per se'. But I hope there can be some productive co-existence. After all, some IIIF communities do venture into modeling their objects of interest, albeit with a presentation focus of course.
I also see this in what Ben has shared on Slack: https://github.com/benwbrum/fromthepage/wiki/FromThePage-Support-for-the-IIIF-Presentation-API-and-Web-Annotations-(draft)
where the sc:painting motivation is in the 'deprecated' pattern and he has added a big 'TODO' for the motivation in the 'proposed' one :-)
At least, even if we keep sc:painting or its successor for compatibility with 'basic' presentation purposes, we could have a second motivation that represents the transcription nature of the annotation. Which means we'd diverge with the Web Annotation official recommendation ("There SHOULD be exactly 1 motivation for each Annotation, and MAY be 0 or more than 1." at https://www.w3.org/TR/annotation-model/#motivation-and-purpose). Your opinion would be very welcome on this too :-)
Cheers,
Antoine
On 20/07/17 22:41, Robert Sanderson wrote:
>
> The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
>
> So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.
>
> There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
>
> For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
>
> As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
>
> HTH
>
> Rob
>
>
> [1] https://github.com/IIIF/iiif.io/issues/511
>
>
> On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl <mailto:ais...@few.vu.nl>> wrote:
>
> Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
>
> Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
>
> There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
>
> With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
>
> We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
>
> Thanks for the help!
>
> Antoine
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> --- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss+unsub...@googlegroups.com>.
Antoine Isaac
Jul 21, 2017, 11:35:45 AM7/21/17
to iiif-d...@googlegroups.com
Hi Rob, all,
Thanks for the input!
By "re-use" I mean that the statements that represent transcriptions could be used in different contexts (IIIF, Text modeling) with as few changes as possible. Especially, while the target may be specific (the IIIF context expects IIIF resources like canvases) the other components of the annotation could remain the same. Especially the body, and - I was hoping - the motivation.
I reckon it may be tricky to blend a strict 'presentation' perspective with an approach that goes more into modeling the transcription 'per se'. But I hope there can be some productive co-existence. After all, some IIIF communities do venture into modeling their objects of interest, albeit with a presentation focus of course.
Thanks for the input!
By "re-use" I mean that the statements that represent transcriptions could be used in different contexts (IIIF, Text modeling) with as few changes as possible. Especially, while the target may be specific (the IIIF context expects IIIF resources like canvases) the other components of the annotation could remain the same. Especially the body, and - I was hoping - the motivation.
I reckon it may be tricky to blend a strict 'presentation' perspective with an approach that goes more into modeling the transcription 'per se'. But I hope there can be some productive co-existence. After all, some IIIF communities do venture into modeling their objects of interest, albeit with a presentation focus of course.
At least, even if we keep sc:painting or its successor for compatibility with 'basic' presentation purposes, we could have a second motivation that represents the transcription nature of the annotation. Which means we'd diverge with the Web Annotation official recommendation ("There SHOULD be exactly 1 motivation for each Annotation, and MAY be 0 or more than 1." at https://www.w3.org/TR/annotation-model/#motivation-and-purpose). Your opinion would be very welcome on this too :-)
Cheers,
Antoine
On 20/07/17 22:41, Robert Sanderson wrote:
>
Cheers,
Antoine
On 20/07/17 22:41, Robert Sanderson wrote:
>
> The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
>
> So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.
>
> There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
>
> For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
>
> As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
>
> HTH
>
> Rob
>
>
> [1] https://github.com/IIIF/iiif.io/issues/511
>
>
> On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl <mailto:ais...@few.vu.nl>> wrote:
>
> Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
>
> Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
>
> There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
>
> With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
>
> We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
>
> Thanks for the help!
>
> Antoine
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
>
> So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.
>
> There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
>
> For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
>
> As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
>
> HTH
>
> Rob
>
>
> [1] https://github.com/IIIF/iiif.io/issues/511
>
>
> On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl <mailto:ais...@few.vu.nl>> wrote:
>
> Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
>
> Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
>
> There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
>
> With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
>
> We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
>
> Thanks for the help!
>
> Antoine
>
> --
> --- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
Ben Brumfield
Jul 21, 2017, 11:45:29 AM7/21/17
to IIIF Discuss, hugo.ma...@europeana.eu, sergiu...@ait.ac.at
I'm okay with sc:painting as a hard-wired motivation value for text representing an image, but am wondering if there are other commonly-used properties which can be used to express a transcription versus a translation? Is purpose also similarly restricted?
Regarding diplomatic vs. normalized transcriptions, could that be a use case for oa:choice, as described at http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources?
Ben
Regarding diplomatic vs. normalized transcriptions, could that be a use case for oa:choice, as described at http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources?
Ben
On Thursday, July 20, 2017 at 3:41:28 PM UTC-5, Rob Sanderson wrote:
The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.HTHRob
On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl> wrote:
Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
Thanks for the help!
Antoine
--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
--- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Shaun Ellis
Jul 21, 2017, 2:04:43 PM7/21/17
to iiif-d...@googlegroups.com
I agree with Ben that there must be a way to tell the difference between transcriptions and translations. The following use case requires such a distinction, and has to do with presenting information to the user (not modeling):
As a transcriber, I want to know (at a glance) which pages of a diary have already been transcribed so that I can select a page to work on.
If the annotation store contains both transcriptions and translations, how can I create an interface that will fulfill this use case?
Another use case that requires such a distinction would be if I need a screen reader or text-to-speech to read the diary. How do I know if I'm reading a transcription or a translation?
And on that note, in the interest of accessibility, I propose the next version of the Presentation API follow the example of the Web Annotation spec by making the association of language with strings a SHOULD rather than a MAY in section 4.3.
-Shaun
--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
---
You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss+unsubscribe@googlegroups.com.
Patrick Cuba
Jul 21, 2017, 2:57:12 PM7/21/17
to iiif-d...@googlegroups.com
This is also complicated for us (OngCDH, TPEN transcription tool next version development). One possible solution we've kicked around is leaving the annotations as `sc:painting` since they are in the list to draw themselves on the canvas, but to motivate the resource from which they are drawn (either the `chars` of each line or page or document at once from which they are `#selected`) with something to clearly state that the content is translation, transcription, commentary, etc.
I agree with investigating several SHOULD/MAY/MUST language instances, but I can also see that for something like UV or Mirador to work, the annotations drawn onto the Canvas by default MUST be recognizable in a clear way—it may not be that anything with a `selector` is relevant for display in place.
To the use cases, the type/class/motivation of the annotations (or the list that aggregates them) should be enough to determine which pages are vacant in the diary and is only barely touching the IIIF spec itself. The text-to-speech would be best served by a full page text service of some type which would not cause issues in determining the proper order of annotations that have word or line granularity and which alternate between main content, glosses, and marginalia.
I believe strongly that a good cookbook for IIIF should offer a best practice for this type of thing, but it may not be something IIIF itself can address directly. Mirador (et al.) support is a separate concern.
Robert Sanderson
Jul 21, 2017, 4:27:14 PM7/21/17
to iiif-d...@googlegroups.com
On Fri, Jul 21, 2017 at 11:04 AM, Shaun Ellis <sh...@sdellis.com> wrote:
As a transcriber, I want to know (at a glance) which pages of a diary have already been transcribed so that I can select a page to work on.
If the annotation store contains both transcriptions and translations, how can I create an interface that will fulfill this use case?
The transcriptions and translations and editions and ... can (and must) all live in separate Layers, which have labels. In this way, we fulfill the presentation case without getting into the semantics of textual analysis, which has been worked on in the TEI and other scholarly communities for decades.
Another use case that requires such a distinction would be if I need a screen reader or text-to-speech to read the diary. How do I know if I'm reading a transcription or a translation?
As above. Even better, the screen reader doesn't need to know the distinction, it just does what it always does and reads the label of the layers available to present to the user :) Then it's infinitely extensible to any collection of annotations, not just the transcription/edition/translation specific case.
And on that note, in the interest of accessibility, I propose the next version of the Presentation API follow the example of the Web Annotation spec by making the association of language with strings a SHOULD rather than a MAY in section 4.3.
Yes, it needs to if we're to use the much more developer-friendly language maps, as per the examples and as per issue 755 [1]
Rob
Shaun Ellis
Jul 21, 2017, 5:30:14 PM7/21/17
to iiif-d...@googlegroups.com
Thanks, Rob. With your explanation, I agree with your solution for the screen reader use case.
However, I worry about the interoperability of using labels to distinguish transcription Layers from others. For example, let's say there is a Manifest that contains Canvases aggregated from various institutions (i.e., correspondence from multiple archives). You may have existing transcriptions that are in multiple annotation stores, each of which labels the transcription Layer differently (i.e., "full-text" vs "transcriptions"). In such a case, how does an interface like this one know whether or not to put a "started" or "not started" label on the thumbnail of each page?
-Shaun
David Newbury
Jul 22, 2017, 12:08:29 PM7/22/17
to iiif-d...@googlegroups.com
I would think that it's out-of-scope to try and define and enumerate every possible use case (and required flag or metadata) that an application might want to perform on a IIIF manifest. For instance, even for the transcription use case, a hypothetical crowd-sourced transcription system might need to record variant transcriptions and a measure of consensus. Another might need to record approval status, to distinguish between reviewed and transcribed. These are all valid use cases, but certainly not an exhaustive list!
If there's an interest in interoperable transcription applications, there's definitely a need for some level of standardization in this kind of metadata. However, it feels like it's a specific use case and should be defined as an extension or a cookbook pattern, not as part of a global IIIF specification. I feel the same way about some of the discovery conversations, as well as computational analysis of images (my own personal interest.) If IIIF needs to be extended to support generalizable patterns such as this, that seems appropriate, but figuring out patterns for accommodating every potential use case seems like a way to add unsupportable complexity.
That being said, I would love to be involved in conversations about what a more formalized pattern for such extensions might be, if anyone else is interested.
Antoine Isaac
Jul 24, 2017, 8:28:28 AM7/24/17
to iiif-d...@googlegroups.com
Hi everyone,
I would also agree that it's out of scope for the 'core' IIIF spec to specify every possible type of annotation.
A cookbook would be fine.
That said the core IIIF spec could also be impacted if the practices it encourages are consider sub-optimal by the community. There are two aspects that would bother me...
(but first a caveat: I have to say that I've not given all this a lot of thinking! In particular, my worries are based on the assumption that from a perspective focused on the Web Annotation model, it is natural to represent a transcription as an annotation with a motivation of transcription)
1. sc:painting used as motivation somehow 'fills the slot' for other, more 'business' motivations on the annotations, like saying that an annotation is motivated for transcription purposes. The Web Annotation standard recognizes that an annotation MAY have several motivations, but it definitely says that it SHOULD have one (https://www.w3.org/TR/annotation-model/#motivation-and-purpose). So I'd say that cookbook that says that transcription should be represented as a motivation on the Annotation, would probably be helped if IIIF Presentation API makes it explicit that its use of sc:painting shouldn't discourage implementers to use other motivations.
2. Representing translations/transcriptions/etc as different Layers only make sense in a IIIF silo. But in a context where annotation resources are re-used from one application to another, the annotations in a 'transcription Layer' would still have a 'transcription motivation', making a part of the Layer definition a bit redundant. Unless the idea is that annotations coming from a different environment would be 'forked' (new URIs, different statements) when the IIIF manifest is created, to get new, 'pure IIIF' anotations (i.e. no more 'transcription motivation on them, only sc:painting)?
Cheers,
Antoine
On 22/07/17 18:08, David Newbury wrote:
> I would think that it's out-of-scope to try and define and enumerate every possible use case (and required flag or metadata) that an application might want to perform on a IIIF manifest. For instance, even for the transcription use case, a hypothetical crowd-sourced transcription system might need to record variant transcriptions and a measure of consensus. Another might need to record approval status, to distinguish between reviewed and transcribed. These are all valid use cases, but certainly not an exhaustive list!
>
> If there's an interest in interoperable transcription applications, there's definitely a need for some level of standardization in this kind of metadata. However, it feels like it's a specific use case and should be defined as an extension or a cookbook pattern, not as part of a global IIIF specification. I feel the same way about some of the discovery conversations, as well as computational analysis of images (my own personal interest.) If IIIF needs to be extended to support generalizable patterns such as this, that seems appropriate, but figuring out patterns for accommodating every potential use case seems like a way to add unsupportable complexity.
>
> That being said, I would /love /to be involved in conversations about what a more formalized pattern for such extensions might be, if anyone else is interested.
> e. david....@gmail.com <mailto:david....@gmail.com>
>
> -Shaun
>
> On Fri, Jul 21, 2017 at 4:27 PM, Robert Sanderson <azar...@gmail.com <mailto:azar...@gmail.com>> wrote:
>
>
>
> On Fri, Jul 21, 2017 at 11:04 AM, Shaun Ellis <sh...@sdellis.com <mailto:sh...@sdellis.com>> wrote:
>
> /As a transcriber, I want to know (at a glance) which pages of a diary have already been transcribed so that I can select a page to work on./
>
>
>
>
> -Shaun
>
> On Fri, Jul 21, 2017 at 11:45 AM, Ben Brumfield <benw...@gmail.com <mailto:benw...@gmail.com>> wrote:
>
> I'm okay with *sc:painting* as a hard-wired motivation value for text representing an image, but am wondering if there are other commonly-used properties which can be used to express a transcription versus a translation? Is *purpose* also similarly restricted?
>
> Regarding diplomatic vs. normalized transcriptions, could that be a use case for *oa:choice*, as described at http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources <http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources>?
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
I would also agree that it's out of scope for the 'core' IIIF spec to specify every possible type of annotation.
A cookbook would be fine.
That said the core IIIF spec could also be impacted if the practices it encourages are consider sub-optimal by the community. There are two aspects that would bother me...
(but first a caveat: I have to say that I've not given all this a lot of thinking! In particular, my worries are based on the assumption that from a perspective focused on the Web Annotation model, it is natural to represent a transcription as an annotation with a motivation of transcription)
1. sc:painting used as motivation somehow 'fills the slot' for other, more 'business' motivations on the annotations, like saying that an annotation is motivated for transcription purposes. The Web Annotation standard recognizes that an annotation MAY have several motivations, but it definitely says that it SHOULD have one (https://www.w3.org/TR/annotation-model/#motivation-and-purpose). So I'd say that cookbook that says that transcription should be represented as a motivation on the Annotation, would probably be helped if IIIF Presentation API makes it explicit that its use of sc:painting shouldn't discourage implementers to use other motivations.
2. Representing translations/transcriptions/etc as different Layers only make sense in a IIIF silo. But in a context where annotation resources are re-used from one application to another, the annotations in a 'transcription Layer' would still have a 'transcription motivation', making a part of the Layer definition a bit redundant. Unless the idea is that annotations coming from a different environment would be 'forked' (new URIs, different statements) when the IIIF manifest is created, to get new, 'pure IIIF' anotations (i.e. no more 'transcription motivation on them, only sc:painting)?
Cheers,
Antoine
On 22/07/17 18:08, David Newbury wrote:
> I would think that it's out-of-scope to try and define and enumerate every possible use case (and required flag or metadata) that an application might want to perform on a IIIF manifest. For instance, even for the transcription use case, a hypothetical crowd-sourced transcription system might need to record variant transcriptions and a measure of consensus. Another might need to record approval status, to distinguish between reviewed and transcribed. These are all valid use cases, but certainly not an exhaustive list!
>
> If there's an interest in interoperable transcription applications, there's definitely a need for some level of standardization in this kind of metadata. However, it feels like it's a specific use case and should be defined as an extension or a cookbook pattern, not as part of a global IIIF specification. I feel the same way about some of the discovery conversations, as well as computational analysis of images (my own personal interest.) If IIIF needs to be extended to support generalizable patterns such as this, that seems appropriate, but figuring out patterns for accommodating every potential use case seems like a way to add unsupportable complexity.
>
> e. david....@gmail.com <mailto:david....@gmail.com>
>
> On Fri, Jul 21, 2017 at 2:30 PM, Shaun Ellis <sh...@sdellis.com <mailto:sh...@sdellis.com>> wrote:
>
> Thanks, Rob. With your explanation, I agree with your solution for the screen reader use case.
>
> However, I worry about the interoperability of using labels to distinguish transcription Layers from others. For example, let's say there is a Manifest that contains Canvases aggregated from various institutions (i.e., correspondence from multiple archives). You may have existing transcriptions that are in multiple annotation stores, each of which labels the transcription Layer differently (i.e., "full-text" vs "transcriptions"). In such a case, how does an interface like this one <http://diyhistory.lib.uiowa.edu/items/show/3112> know whether or not to put a "started" or "not started" label on the thumbnail of each page?
> On Fri, Jul 21, 2017 at 2:30 PM, Shaun Ellis <sh...@sdellis.com <mailto:sh...@sdellis.com>> wrote:
>
> Thanks, Rob. With your explanation, I agree with your solution for the screen reader use case.
>
>
> -Shaun
>
> On Fri, Jul 21, 2017 at 4:27 PM, Robert Sanderson <azar...@gmail.com <mailto:azar...@gmail.com>> wrote:
>
>
>
> On Fri, Jul 21, 2017 at 11:04 AM, Shaun Ellis <sh...@sdellis.com <mailto:sh...@sdellis.com>> wrote:
>
> /As a transcriber, I want to know (at a glance) which pages of a diary have already been transcribed so that I can select a page to work on./
>
>
> If the annotation store contains both transcriptions and translations, how can I create an interface that will fulfill this use case?
>
>
> The transcriptions and translations and editions and ... can (and must) all live in separate Layers, which have labels. In this way, we fulfill the presentation case without getting into the semantics of textual analysis, which has been worked on in the TEI and other scholarly communities for decades.
>
>
>
> Another use case that requires such a distinction would be if I need a screen reader or text-to-speech to read the diary. How do I know if I'm reading a transcription or a translation?
>
>
> As above. Even better, the screen reader doesn't need to know the distinction, it just does what it always does and reads the label of the layers available to present to the user :) Then it's infinitely extensible to any collection of annotations, not just the transcription/edition/translation specific case.
>
>
> And on that note, in the interest of accessibility, I propose the next version of the Presentation API follow the example of the Web Annotation spec by making the association of language with strings a SHOULD rather than a MAY in section 4.3 <http://iiif.io/api/presentation/2.1/#language-of-property-values>.
>
> If the annotation store contains both transcriptions and translations, how can I create an interface that will fulfill this use case?
>
>
> The transcriptions and translations and editions and ... can (and must) all live in separate Layers, which have labels. In this way, we fulfill the presentation case without getting into the semantics of textual analysis, which has been worked on in the TEI and other scholarly communities for decades.
>
>
>
> Another use case that requires such a distinction would be if I need a screen reader or text-to-speech to read the diary. How do I know if I'm reading a transcription or a translation?
>
>
> As above. Even better, the screen reader doesn't need to know the distinction, it just does what it always does and reads the label of the layers available to present to the user :) Then it's infinitely extensible to any collection of annotations, not just the transcription/edition/translation specific case.
>
>
>
>
> Yes, it needs to if we're to use the much more developer-friendly language maps, as per the examples and as per issue 755 [1]
>
> Rob
>
> [1] https://github.com/IIIF/iiif.io/issues/755 <https://github.com/IIIF/iiif.io/issues/755>
>
> Yes, it needs to if we're to use the much more developer-friendly language maps, as per the examples and as per issue 755 [1]
>
> Rob
>
>
>
>
>
> -Shaun
>
> On Fri, Jul 21, 2017 at 11:45 AM, Ben Brumfield <benw...@gmail.com <mailto:benw...@gmail.com>> wrote:
>
> I'm okay with *sc:painting* as a hard-wired motivation value for text representing an image, but am wondering if there are other commonly-used properties which can be used to express a transcription versus a translation? Is *purpose* also similarly restricted?
>
> Regarding diplomatic vs. normalized transcriptions, could that be a use case for *oa:choice*, as described at http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources <http://iiif.io/api/presentation/2.1/#choice-of-alternative-resources>?
>
> Ben
>
> On Thursday, July 20, 2017 at 3:41:28 PM UTC-5, Rob Sanderson wrote:
>
>
> The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
>
> So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content <http://iiif.io/api/presentation/2.1/#embedded-content> is how we would do transcription in the Presentation API with Annotations.
> Ben
>
> On Thursday, July 20, 2017 at 3:41:28 PM UTC-5, Rob Sanderson wrote:
>
>
> The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
>
>
> There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
>
> For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
>
> As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
>
> HTH
>
> Rob
>
>
> [1] https://github.com/IIIF/iiif.io/issues/511 <https://github.com/IIIF/iiif.io/issues/511>
> There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
>
> For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
>
> As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
>
> HTH
>
> Rob
>
>
>
>
> On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl> wrote:
>
> Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
>
> Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
>
> There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
>
> With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
>
> We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
>
> Thanks for the help!
>
> Antoine
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
>
> On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl> wrote:
>
> Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
>
> Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
>
> There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
>
> With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
>
> We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
>
> Thanks for the help!
>
> Antoine
>
> --
> --- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
>
>
>
> --
> Rob Sanderson
> Semantic Architect
> The Getty Trust
> Los Angeles, CA 90049
>
> --
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <mailto:iiif-d...@googlegroups.com>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss%2Bunsu...@googlegroups.com>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
John Howard
Jul 24, 2017, 11:40:13 AM7/24/17
to IIIF Discuss, hugo.ma...@europeana.eu, sergiu...@ait.ac.at
Hi All,
Responding to Rob's early posting in this thread to note a particular transcription use case.
At one point I was responsible for teaching paleography and would assign transcription work based on facsimiles to students; this included text transcription, but also transcription into modern music notation of music notated in medieval/renaissance notation systems, so there were both textual transcriptions and renderings of musical text into modern notation. Even if we confine the use case to textual transcription, one has to reckon with multiple transcriptions by students of the same text--not permanent contributions to the scholarly record, but ephemeral course-work undertaken for learning purposes. So in this scenario we have multiple manifestations of transcriptions that relate directly to the image on the canvas; I'd see these as layers. (Ideally, this use case would also be supported by some mechanism for authenticating/authorising users/groups, similar to what hypothes.is supports.)
I've been working on a Mirador plugin that would link users of the local resource with FromThePage; ultimately we would wish to harvest completed transcriptions from the transcription site and ingest them into our repository, linking them to the original image/document sources. Note that the TEI object harvested makes references to IIIF image resources; an example showing this with both a transcription and translation is seen at https://github.com/IIIF/iiif-stories/issues/100. It would not be difficult to extract individual page transcriptions and disseminate in annotation lists (associated with the entire canvas). One of the challenges with this integration generally is deciding how to make a Mirador/IIIF user aware of the potentially many manifestations of a page or document transcription without having to transition into a completely different environment.
John
On Thursday, 20 July 2017 21:41:28 UTC+1, Rob Sanderson wrote:
The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content is how we would do transcription in the Presentation API with Annotations.There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.HTHRob
On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl> wrote:
Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
Thanks for the help!
Antoine
--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
--- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Antoine Isaac
Jul 25, 2017, 5:08:27 PM7/25/17
to iiif-d...@googlegroups.com
Dear Masahide,
Coming back to your example a bit late... this looks like a very interesting case. I'm wondering: how do you know that the annotations [2] are transcriptions as opposed to any other type of annotation? Or is it that you don't need to know this?
Cheers,
Antoine
On 21/07/17 13:27, KANZAKI Masahide wrote:
> Hello Antoine, all
>
> I have an experiment that connects IIIF images and TEI/XML transcription in Bodleian First Folio [1].
>
> It does not embed link information in the manifest file. Rather, it uses separate Web Annotation description where the targets are images and bodies are TEI pages [2]. Although the targets are not canvases, it could be easily translated into IIIF AnnotationList.
>
> There was a discussion in TEI mailing list on how to relate TEI and corresponding facsimile (image) [3].
>
> hope this would be relevant.
>
> cheers,
>
> [1] http://www.kanzaki.com/works/ld/firstfolio
> [2] http://www.kanzaki.com/works/ld/bodleian/hamlet.jsonld
> [3] https://listserv.brown.edu/archives/cgi-bin/wa?A1=ind1706&L=tei-l#8
>
> 2017年7月21日金曜日 18時13分17秒 UTC+9 Antoine Isaac:
>
> Hi Rob, all,
>
> Thanks for the input!
>
> By "re-use" I mean that the statements that represent transcriptions could be used in different contexts (IIIF, Text modeling) with as few changes as possible. Especially, while the target may be specific (the IIIF context expects IIIF resources like canvases) the other components of the annotation could remain the same. Especially the body, and - I was hoping - the motivation.
>
> I reckon it may be tricky to blend a strict 'presentation' perspective with an approach that goes more into modeling the transcription 'per se'. But I hope there can be some productive co-existence. After all, some IIIF communities do venture into modeling their objects of interest, albeit with a presentation focus of course.
>
> I also see this in what Ben has shared on Slack: https://github.com/benwbrum/fromthepage/wiki/FromThePage-Support-for-the-IIIF-Presentation-API-and-Web-Annotations-(draft) <https://github.com/benwbrum/fromthepage/wiki/FromThePage-Support-for-the-IIIF-Presentation-API-and-Web-Annotations-(draft)>
> > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout> <https://groups.google.com/d/optout <https://groups.google.com/d/optout>>.
Coming back to your example a bit late... this looks like a very interesting case. I'm wondering: how do you know that the annotations [2] are transcriptions as opposed to any other type of annotation? Or is it that you don't need to know this?
Cheers,
Antoine
On 21/07/17 13:27, KANZAKI Masahide wrote:
> Hello Antoine, all
>
> I have an experiment that connects IIIF images and TEI/XML transcription in Bodleian First Folio [1].
>
> It does not embed link information in the manifest file. Rather, it uses separate Web Annotation description where the targets are images and bodies are TEI pages [2]. Although the targets are not canvases, it could be easily translated into IIIF AnnotationList.
>
> There was a discussion in TEI mailing list on how to relate TEI and corresponding facsimile (image) [3].
>
> hope this would be relevant.
>
> cheers,
>
> [1] http://www.kanzaki.com/works/ld/firstfolio
> [2] http://www.kanzaki.com/works/ld/bodleian/hamlet.jsonld
> [3] https://listserv.brown.edu/archives/cgi-bin/wa?A1=ind1706&L=tei-l#8
>
> 2017年7月21日金曜日 18時13分17秒 UTC+9 Antoine Isaac:
>
> Hi Rob, all,
>
> Thanks for the input!
>
> By "re-use" I mean that the statements that represent transcriptions could be used in different contexts (IIIF, Text modeling) with as few changes as possible. Especially, while the target may be specific (the IIIF context expects IIIF resources like canvases) the other components of the annotation could remain the same. Especially the body, and - I was hoping - the motivation.
>
> I reckon it may be tricky to blend a strict 'presentation' perspective with an approach that goes more into modeling the transcription 'per se'. But I hope there can be some productive co-existence. After all, some IIIF communities do venture into modeling their objects of interest, albeit with a presentation focus of course.
>
> where the sc:painting motivation is in the 'deprecated' pattern and he has added a big 'TODO' for the motivation in the 'proposed' one :-)
>
> At least, even if we keep sc:painting or its successor for compatibility with 'basic' presentation purposes, we could have a second motivation that represents the transcription nature of the annotation. Which means we'd diverge with the Web Annotation official recommendation ("There SHOULD be exactly 1 motivation for each Annotation, and MAY be 0 or more than 1." at https://www.w3.org/TR/annotation-model/#motivation-and-purpose <https://www.w3.org/TR/annotation-model/#motivation-and-purpose>). Your opinion would be very welcome on this too :-)
>
>
> Cheers,
>
> Antoine
>
> On 20/07/17 22:41, Robert Sanderson wrote:
> >
> > The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
> >
> > So, other than the obvious updates for WA rather than Open Annotation, the example in http://iiif.io/api/presentation/2.1/#embedded-content <http://iiif.io/api/presentation/2.1/#embedded-content> is how we would do transcription in the Presentation API with Annotations.
> Cheers,
>
> Antoine
>
> On 20/07/17 22:41, Robert Sanderson wrote:
> >
> > The rationale for sc:painting is that the transcription is a representation of the content of the canvas, as opposed to an annotation /about/ the canvas (which would use the regular Web Annotation motivations).
> >
> >
> > There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
> >
> > For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
> >
> > As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
> >
> > HTH
> >
> > Rob
> >
> >
> > [1] https://github.com/IIIF/iiif.io/issues/511 <https://github.com/IIIF/iiif.io/issues/511>
> > There's an open issue [1] about whether it is IIIF's issue to represent /semantic/ distinctions such as between diplomatic transcriptions, multi-witness editions, translations and so forth. Keeping in mind that the Presentation API is about presenting information to the user, not about semantic textual modeling, I feel like the right approach to coming to a conclusion for this is to look at how user agents would present transcriptions versus editions differently, if at all.
> >
> > For that, it would be valuable to have folks involved who have actively built or are actively building such rendering tools, as well as the end users.
> >
> > As to reuse outside of the Presentation API ... what would be re-used, and how? The transcription targets the Canvas (of course) so it's somewhat tied together.
> >
> > HTH
> >
> > Rob
> >
> >
> >
> >
> > On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl <javascript:> <mailto:ais...@few.vu.nl <javascript:>>> wrote:
> >
> > Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
> >
> > Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
> >
> > There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
> >
> > With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
> >
> > We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
> >
> > Thanks for the help!
> >
> > Antoine
> >
> > --
> > -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <javascript:> <mailto:iiif-d...@googlegroups.com <javascript:>>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <javascript:> <mailto:iiif-discuss%2Bunsu...@googlegroups.com <javascript:>>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en> <https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>>
> >
> > On Thu, Jul 20, 2017 at 10:58 AM, Antoine Isaac <ais...@few.vu.nl <javascript:> <mailto:ais...@few.vu.nl <javascript:>>> wrote:
> >
> > Hi everyone, (with a special heads-up to friends in the Manuscript, Text Granularity and Newspapers communities)
> >
> > Europeana has to create a model for transcriptions. We're keen on using Web Annotations. Are there cases in the IIIF community where Web Annotations has been used to represent transcriptions in the Presentation API, with pointers that we could look at directly?
> >
> > There are a couple of thread from 2015 on this list, but of course I'm interested in applications that would have considered the latest versions of Web Annotations and the Presentation API.
> >
> > With all the case of manuscripts around (or even OCR, which I assume is a quite similar pattern) I'd be surprised if it wasn't the case, but I don't have the time to dig it up...
> >
> > We're especially interested in re-using patterns for Web Annotation Motivations and Scopes. The old example I could find use sc:painting as the motivation for transcriptions, and I'm wondering whether this would be the best way to have annotations that could also be re-used outside the Presentation API context (which is what I'm also interested in).
> >
> > Thanks for the help!
> >
> > Antoine
> >
> > --
> > --- You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> > To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <javascript:> <mailto:iiif-discuss%2Bunsu...@googlegroups.com <javascript:>>.
> > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout> <https://groups.google.com/d/optout <https://groups.google.com/d/optout>>.
> >
> >
> >
> >
> > --
> > Rob Sanderson
> > Semantic Architect
> > The Getty Trust
> > Los Angeles, CA 90049
> >
> > --
> > -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com <javascript:>. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com <javascript:>. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en <https://groups.google.com/d/forum/iiif-discuss?hl=en>
> >
> >
> >
> > --
> > Rob Sanderson
> > Semantic Architect
> > The Getty Trust
> > Los Angeles, CA 90049
> >
> > --
> > ---
> > You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> > To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <javascript:> <mailto:iiif-discuss...@googlegroups.com <javascript:>>.
> > You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
> --
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss...@googlegroups.com>.
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
KANZAKI Masahide
Jul 26, 2017, 9:50:45 AM7/26/17
to IIIF Discuss
Hello Antoine,
Well, each annotation only says that the specific range of the TEI/XML is related to the target image (canvas). This tool tries to find occurrences of a word in TEI/XML, determines the page ranges of found positions, then presents the text of selected range (body) and the corresponding target.
Therefore, the tool doesn't need to know what is the relation between TEI and IIIF (actually, it can be Japanese translation of Hamlet), although its intention is to present transcription.
Of course, motivation property will be useful for some other tools to differentiate the presentation according to the relation.
I guess it would be another way to declare source level relation (TEI and manifest) and describe each target as part of the manifest.
cheers,
2017年7月26日水曜日 6時08分27秒 UTC+9 Antoine Isaac:
> > To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <javascript:> <mailto:iiif-discuss+unsub...@googlegroups.com <javascript:>>.
> > For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
> --
> -- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com <mailto:iiif-discuss+unsub...@googlegroups.com>.
Reply all
Reply to author
Forward
0 new messages