2016-02-03 @ 1300 EST — Meeting notes
- Bill Branan
Date
Link to join Hangout: https://plus.google.com/hangouts/_/calendar/dmlsbGVyZWFsQGdtYWlsLmNvbQ.43lla0mm962pt9c0apfa3n09p4
Attendees
- Juliet Hardesty
- Justin Simpson
- Karen Cariani
- Michael Angeletti (Stanford)
- Brian Davis
- Rebecca Fraimow
Goals
First goal is to come up with a recommendation for how to extend the current Hydra Metadata Application Profile, to include a base profile for technical metadata about av file formats.
Discussion items
| Time | Item | Who | Notes |
|---|---|---|---|
| 10 min | Recap from AMIA face-to-face meeting, short-term goals for this group | ||
| 10 min | Hear from Technical Metadata Working Group members about how the approached their tech MD profile, and how we might proceed with AV-specific work | ||
| 15 min | Discuss technical metadata characterization workflows. What are folks short-term plans re: RDF/Fedora 4? | ||
| 15 min | Discussion of initial use cases for audio/video technical metadata in RDF |
The group discussed the work that is going on at Indiana University and WGBH, in the HydraDAM project. They have done work to map the output of fits/ffprobe/exif and pbcore to rdf properties. They have also started to figure out how they want to model av content using PCDM.
IU is not storing av master files in Fedora, same thing at WGBH. The digitization work is done by a vendor, it is simply too slow to do within Hydra/Fedora. The PCDM model for this workflow is documented here: https://docs.google.com/drawings/d/13Ir6VVweuVHNfLX2NhW6790TpZXU4eHRUaHyhIrKlC0/edit
We discussed the Preforma Project, and the work that MediaArea has done to provide ebucore xml output and mediatrace xml out from their characterization tools (mediainfo, mediaconch).
Rebecca Fraimow reported that there is a pbcore interest group - working on an edubcore/pbcore mapping, and how to map to rdf. She will report back to this group about progress/work there.
First step for this group is to follow the basic approach that was used by the Hydra Technical Metadata Interest Group
- collaborate in a google spreadsheet, to identify a small starting set of metadata properties that are required for specific file format type
- ►►► spreadsheet via Juliet Hardesty documenting properties used by HydraDAM2 sites: https://docs.google.com/spreadsheets/d/12W5aqE68xZh5SW7Vzr04uzR8-rMKE3r4_UR7M-RTXCo/edit#gid=0 ...I've left a version arranged by tool for now, but anyone should feel free to edit
- evaluate existing ontologies (nfo, ebucore, pbcore, sweetjpl, premis, others?) to find best candidates for each required property
- explicitly tie each set of recommended properties to one of the File Format Types defined in the pcdm-ext file format type name space: https://github.com/duraspace/pcdm/blob/master/pcdm-ext/file-format-types.rdf
- review the spreadsheet at our next monthly meeting, and formalize the recommendations in a wiki page similar to the current Hyrda MAP: Technical Metadata Application Profile
Next steps include
- making specific recommendations about how to map the output of specific characterization tools to each of these recommended properties. Juliet Hardesty will provide a start on this, IU/WGBH have already done some of this work for ffprobe and fits (also exif,pbcore).
- creating some real world examples of what a set of av files look like when described using PCDM, with the base Hydra MAP and these recommended extensions included.
Action items
- villereal will create a google spreadsheet and link it to this wiki
- Juliet Hardesty will add content from IU/WGBH work on HydraDAM, to that spreadsheet
- Rest of the group will review the spreadsheet and add their ideas for what technical metadata properties should be added for each file format type