Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • EDM

...

...

    • data attribution, provenance and licensing
    • expressing authority/trust for links and identifiers (Maybe wiser to do a TF on data attribution first)
    • approximate geo coordinates, or places that cover an area (can be part of a TF on validating EDM template for Places)
    • data versioning
    • interoperability with schema.org and EU ISA vocabularies

    • experiments: best practices/case studies for mapping to EDM
    • documentation: richer publishing of EDM (e.g. HTML documents for Linked Data redirection)
    • infrastructure for EDM data (semantic repositories and others)

Some of these TF ideas are inspired by an Open Issues document. The items there belong to quite different levels (using the numbers from the doc):

  • 1-6 are little improvements or clarifications to make. I'm not sure there should be a task force for them. Maybe that's more work for our EDM "expert group"--but still, we need to activate it.
  • 7-9 were quite clear candidate for task forces. Though for 7 we may want to focus on one category (Agent, Event) at a time, as I've suggested above for Places.
  • 10 is a quite complex set of issues which involve mapping practices and possible extensions. I think it should be a longer-run effort between the EDM expert group and EuropeanaLibraries.
  • Data Quality-related:
    • strategy: selecting relevant data resources for enrichment (items, events, periods, places in history
    • tools for semantic extraction from object metadata (name extraction, topic extraction), object clustering (loose FRBRization), alignment of reference contextual resources (thesauri etc.)
    • survey: keeping track of enrichment of Europeana metadata
  • Techniques for advanced search
    • clustering and/or ranking of (objects) search results
    • recommendation/personalization (cf Karl's experiments)
    • timeline and geo functions (cf Pelagios experiment)
    • content-based search: similarity-based search, full-text search (cf AIT's work in DSI-2 on image similarity and DSI-1 on audio features)
  • User-contributed data
    • includes error fixing (original metadata on enrichment), tagging, adding new fields on a record