AmitMaitraSuggestions

The purpose of the DRM pilot should be to evaluate several different options, document the pros and cons of each approach, and recommend one or more prototype developmernt for the best approach(es) that we can test in a development lab with live databases connected to the prototype(s).    (2YTP)

Further, as part of an integrated test plan, we should address several issues:    (2YTQ)

• Collect data based on an agreed upon DRM Abstract Model and XML Schema, and then clearly document:    (2YTR)
  • 1. Where and how we store the collected data    (2YTS)
  • 2. If we make use of any central registry at the agency level or    (2YTT)
  • 3. Do we identify local registry which rolls up to a regional and on to a central agency registry    (2YTU)
  • 4. What does the registry landscape look like    (2YTV)
  • 5. Who does the parsing and harmonization at the local, regional, and central registry    (2YTW)
  • 6. If there should be a data panel    (2YTX)
  • 7. Who authorizes the data panel    (2YTY)
  • 8. How are the members of the data panel chosen    (2YTZ)
  • 9. What authority does it have for adjudication of data definitions/standardization    (2YU0)
  • 10. Taxonomy - same issues    (2YU1)

• Net-Centric Data Strategy opposes any top-down approach to taxonomy or data standardization efforts. So we should also be evaluating an alternative approach to creating:    (2YU2)
  • 1. Metadata (1M), meta-meta (2M) data, and meta-meta-meta (3M) data that:    (2YU3)
    • 1. Automatically create taxonomy for a particular community of practice in real-time based on a bottom-up approach    (2YU4)
    • 2. Can be captured not only on a pre-defined XML Schema but on any other XML format prepared by an agency    (2YU5)
  • 2. A virtual enterprise data quality awareness - again - in real time    (2YU6)
  • 3. A Federated Data View that could have multiple layers which would:    (2YU7)
    • 1. Correspond to a traditional Logical Data Model (LDM) and Conceptual Data Model (CDM), and, in effect, would    (2YU8)
    • 2. Verify and validate the accuracy of the standard LDM and CDM    (2YU9)
    • 3. Invoke the basic principles of Model Driven Architecture (MDA), thereby facilitating not merely integration but both integration and analysis of data to improve their visibility, accessibility, understandability, usability, and trustworthiness    (2YUA)
    • 4. A data collection process that does not intrude into a data producer's or an information owner's work space in a particular domain.    (2YUB)

• In essence, the data producers and information owners should remain unaffected or unperturbed by any external queries, which is politically more acceptable.    (2YUC)

I think these points should be referenced in the ten-page paper that you are preparing for September 19 submission.    (2YUD)