Goals and Workplan for a Conceptual Model of Cross-DDI Product Objects

Owned by Wendy Thomas
Dec 08, 2022
2 min read

NOTES

Use cases:

  • Variable Cascade including relationships between Lifecycle and CDI

  • Expressing physical content within different products

  • Grouping

  • Name/Label/Description

  • Bridge to CDI to disseminate content - variables (concepts, classifications), data relationships, logical and physical structures

Not a straight mapping project

  • Differentiating packaging from content - focus on core granular objects of interest

  • Automated transformation between products - presentation decisions may be down the road

  • Entry point into object and then what is being layered on by different products to facilitate specific use

Purpose based

  • Mapping to inform what is needed 

  • Support a proliferation of serialization

  • Support transference of content between products

General Needs

  • SKOS provides a minimal set of information (code, label, identifier)

  • Need for Identifiers to support transport between different products (DDI and other)

Workplan

Outputs:

What we mean in terms of basic objects

  • Identify those objects

  • Entry point to describe common objects

  • What is needed to describe differences/commonalities

Overall conceptual model 

  • Differentiate basic objects (content) and packaging/management structures

  • Identify minimal set of information needed for objects

Mapping Examples

  • PRIORITY: Data description - Codebook to CDI and Lifecycle to CDI (outboard critical, inbound needed)

    • CDI - content that points you to some source of metadata (DDI)

      • Address Variables, Instruments, Questions

  • Concept and applied uses

  • Variable Cascade

Draft spreadsheet of objects for discussion 

Note: Package elements can be stripped off without linkage/membership issues. Groups pull together objects that form discrete groups of information about a parent object.

NLD = Name/Label/Description

CDI Documentation

ASIDES from implementation language conversation

Levels of commonality

  • Conceptual level - common core

  • Syntactical level - instances of that class MUST haves

    • Cardinality constraints - not supported but need to relay this information

    • Shackle constraint models for validation

Work Plan going forward - conceptual model

  • Capturing information for producing useful presentations of content

  • Complete the Data Representation spreadsheet for this set of objects - use as a test for capture of content and outputs

  • Create a database / use an existing tool with content for each product covering: 

    • Generic object name

    • Object

    • Type

    • Context (path)

    • For each related product:

      • Target object

      • Path

      • Relationship (SKOS matches)

  • Specific usage notes:

    • Each unique contextual use of an object needs a separate record to differentiate the target path for each product

    • Retain a non-proprietary file (csv, tab delimited, etc.)

  • Complete content for all products

  • Use to generate outputs for various audiences

    • Overall coverage of DDI suite

    • Product centric view from each product

TO DO:

Wendy: 

  • Take current listings of objects in each product and create a tab delimited listing of each product and create initial database records

  • Create recommended list of generic object names (this would be used for the high level conceptual model content)