AVAnnotate Grant Work

[elynema]

Sample AVAnnotate manifests:

• video item (annotation targets contain point selectors for time): https://lgsump.github.io/maya-deren-example/meshes-of-the-afternoon/manifest.json
• audio item (multiple annotation sets and index terms): https://tanyaclement.github.io/znh_jacksonville_1939/s1576-t86-244/manifest.json
• manifest with HTML annotations: https://tanyaclement.github.io/SGW-SI/john-wieners-at-sir-george-williams-university-the-poetry-series-8-october-1966/manifest.json
• no media file in manifest: https://tanyaclement.githuqb.io/harvard1953/august-3-evening-part-one/manifest.json
• multiple Textual Body elements, multiple motivations per annotation, range selectors : https://benwbrum.github.io/znh_rebuild/annotations/s1576-t86-244-canvas-1-songs.json

Tanya's presentation from the IIIF Online Meeting workshop has a nice list of sample manifests, including multi-item manifests.

• Audio: Zora Neale Hurston field recordings at the Library of Congress
• Video: The Lonely Villa (1909) & Suspense (1913)
• Comparison: https://tanyaclement.github.io/shoes/
• Clips: Demo
• Imported Manifest with Captions: UAlberta Writer in Residence Ifeoma Chinwuba and Edmonton Poet Laureate Titilope Sonuga
• Multi-AV Video Event
• Multi-AV Audio Event

Aviary

• item with multiple files (need to support navigating across multiple canvases in Ramp): https://weareavp.aviaryplatform.com/iiif/hm52f7jz70/manifest
• transcripts as textual annotations: https://yalemssa.aviaryplatform.com/iiif/028pc2tq0m/manifest

Structural ranges that cross file boundaries - we don't have an example AVAnnotate or Aviary manifest for this yet

Some things noted that Ramp needs to support:

• multiple motivations in an annotation
• multiple TextualBody elements
• range / point selectors
• transcripts as textual annotations
• multiple annotation sets (annotation pages)
• multiple canvases in a manifest
• structure that spans files/canvases
• commenting and supplementing annotation display (label + value) - need to test to identify other types of annotations might need to support?

[elynema]

Questions for AVAnnotate folks:

• Currently providing filtering based on annotation pages. However, do annotation pages really have semantic value? Should filtering occur by some other mechanism? Sometimes there are tagging labels on the individual annotations, but no label on the annotation page itself. Sometimes there are labels on the annotation page, but no tagging labels on the individual annotations.
• When providing a label to filter annotation sets and to display next to each annotation, do we use the AnnotationPage label (Aviary and AvAnnotate provide) or the textualBody with purpose=tagging label (just AVannotate provides but might be better solution)?
• If there are multiple annotation pages, but no motivation=tagging labels on the annotations themselves, are we supposed to display the annotation page label as a tag on each annotation from that annotation page?
• When an annotation has multiple items in an annotation body, and one of them has a tagging purpose, should we assume that should be displayed as a tag for the annotation text?
• How to handle multiple annotation bodies in a single annotation? Do they all get concatenated into a single annotation? Do we want all textualbodies regardless of purpose or just some?
• Annotations with motivation of commenting (AVAnnotate) and supplementing (Aviary) were called out. Are there other motivation types that we need to support?
• Will textualbodies that have HTML have the format=text/html attached so we know to expect HTML?
  -- From Ben: "I think that's an artifact of us designing the datamodel and then users putting HTML into the app. We probably should change the format to text/html."
• Aviary currently uses motivation=transcribing and motivation=subtitling. Can we ignore annotations with those motivations for now as we expect there to be a new provides property to label them?
• New AVAnnotate manifests have supplementing, commenting, and tagging annotations. For annotations that have both supplementing and commenting motivations, should they be displayed on Transcripts tab, on the Annotations tab, or both? (see annotation page in this example). Should these annotations have a supplementing annotation?

Some examples:

• AVAnnotate: In-line annotation page with label 'Default' but each individual annotation body has a TextualBody with purpose/motivation = tagging. https://tanyaclement.github.io/znh-1939/manifests/manifest.json and https://avannotate.github.io/SuspenseTheLonelyVilla/manifests/manifest.json
• AVAnnotate: External annotation pages with labels. Labels are repeated in the individual annotation body with a TextualBody with purpose/motivation = tagging. https://tanyaclement.github.io/znh_jacksonville_1939/s1576-t86-244/manifest.json
• Aviary: Labels on in-line annotation pages. Each annotation body has a label that is displayed in the annotation, but no TextualBody with purpose/motivation = tagging. https://weareavp.aviaryplatform.com/iiif/hm52f7jz70/manifest

[elynema]

Updated the second manifest linked to a working version. Still having issues with the annotation pages linked in the third manifest.

[elynema]

Should have updated manifests that work now.