Improvements/Avalon

This page will document potential areas of improvement in Archivematica that will allow for better workflows with Avalon.

Project Thesis

Artefactual and Avalon will seek to create:

• support for activities such as file characterization and preservation metadata extraction:
  • digitized and born-digital AV materials.
• Integration with Archivematica or its components into Avalon ingestion workflows to provide:
  • preservation metadata
  • better packaging for long-term digital preservation activities

Use Cases

In scope

Two clear use-cases based on previous discussion (up to 8 March 2018), for which we can focus on and keep scope narrow. These are:

• Avalon first
  • Process in Avalon and then send the package to Archivematica to generate an AIP

Questions

- Do we ignore the Fedora metadata entirely? - Do we transfer Fedora metadata as a digital object, but perform no special conversion mapping - Out-of-scope: Mapping/cross-walk to try and see Fedora metadata mapped into Archivematica

• AM first
  • Collection of assets described in a batch manifest and they are processed and then picked up by Avalon

Two useful phrasings, that could be considered nuances on these, that came from the same meeting:

• As an user of Avalon send the output of a project to a repository and have institutional digital preservation policies applied to that collectionffor long-term preservation
• As a preservation specialist, in a catastrophic scenario, be able to round-trip from AM back into an instance of Avalon**

• • NB Recognised that there is no hard-requirement to round-trip back into Avalon from Archivematica, though it was noted that it would be possible to learn quite a lot.

= Out of scope

Versioning This is a very broad summary, but because of the potential for work to continue being completed on a 'unit' in Avalon, data will become out of date, and there may be a wish to see that updated in Archivematica. This is one example of where data might need updating or to be re-versioned. Others:

• AIP Update following updated in Avalon
• Data supplied as an asset to Avalon/Archivematica, e.g. VTT Subtitle Information, (addition of other languages/corrections, etc.)
• Other cases

What do we need?

- Northwest to work with Sarah to discuss use-cases - Do we need more sample files/batches of data? - Understand the current transfer issue being experienced with the initial setup - In progress.

Questions

METS

• Is it required that the metadata in the batch ingest manifest is represented in the AIP METS file (in dmdSecs for example?)
• Will users have a need to populate the batch ingest manifest with metadata entered into Archivematica?
  • Descriptive metadata?
  • Rights metadata?
• When Avalon is being used to disseminate different quality files, which file(s) should be stored in the AIP?
  • How should mulitple versions be modeled in the METS?
• Should the structure of files be modeled in the METS when structure XML files are included?
  • Or is describing the presence and use of structure files sufficient?
• How should the presence of caption files be noted in the METS?

Exploration of Archivematica

• Progress and first impressions
• Ingest progress
• Exploration of the AIP generated from Avalon Batch Packages
• Immediate observations:
  • Workflow questions
  • Perceived or real gaps
  • Questions Artefactual can answer about what Archivematica is doing/should be doing
• Relationship between Archivematica and Avalon, e.g position in workflow, pre-Avalon? post-Avalon? Something else?