Dataset Metadata: Splinter session


Now that the core models of the Cube family (Meas/Coords) are completed, we return our attention to the next group.

The information in this model is a consolidation of content from several of the early core models (ObsCore, Spectrum, Characterization). The descriptions of many of the elements are derived from the Resource Metadata standard. The idea of this model is to centralize this information for other datamodels to use (Cube, Mango, Spectrum2, TimeSeries).

There are a couple topics which should be discussed so that the DM group can have a clear plan/roadmap going forward:

  • Relation to Provenance Model
    • Much of the content of the Dataset Model can be considered Provenance.
      • Dataset: can be considered an Entity generated by an Activity (software and/or Observation)
      • Party: is equivalent to the Provenance Agent, who serves some role in the history of the Dataset ( Contact, Publisher, etc )
      • several other elements are extractions of the Observation configuration
    • The goal here is to clarify the relation between these models, and how that is represented/described in the Dataset Model.
  • Relation to CAOM2
    • A recurring question has been whether or not CAOM2 can become an IVOA standard.
    • There is significant overlap between that model and Dataset.
      • I believe at one point, I put together a formal comparison of the two.. which I'll try to add here.
    • The goal here is to see if we can make a plan for this proposal and prepare for a smooth migration.
  • Role of Characterization within Dataset
    • (requested by FB)
    • Metadata required to describe datasets must include some overall description of the physical axes of the data (time, spatial, spectral, polarization,etc..). This has been experimented with success in ObsCore
    • Inclusion of a Characterization Package ( set of classes) in DataSet datamodel (as well as in Mango maybe) seems to be useful.
      The issue is that the general IVOA Characterization datamodel (2007) doesn't follow the vodml rules and cannot be integrated "as is" in the DataSet data model.
      This question must be revisited.


To form a better understanding of how these models interrelate and make a plan for this model moving forward.


Please add other discussion topics and/or specific points to cover here.


Edit | Attach | Watch | Print version | History: r4 < r3 < r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r4 - 2023-04-26 - MireilleLouys
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright © 2008-2023 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback