Page History

Discussion Points: 

• Kim described his approach:
  • Kim's approach would eliminate the need for all of the current resolved value sets. 
  • Kim's one OWL file would replace all of these.
  • Can use Protege to find all parent-child (subClassOf) and concept_in_subset (A8) relationships in NCI Thesaurus.
  • Resolving values sets with Kim's OWL file would take about 5 seconds.
• OWL file would be loaded after VS coding schemes are loaded. 
  
  • All 740 value set definitions would be created.
• This would bypass the existing lexevs value set API
• We would use the OWL API to pull data out of the OWL text file.
• We would need extra code generated against the OWL API to accomplish this.
• With this change, the browser will change from the RSV API to the coding scheme API.
• These changes are only for the NCIT. 
• The LexEVS model and API will not change.
• The generated OWL file doesn't have a single header concept.
  • The top node
• Kim confirmed that the value set editors will not have any additional work using this process.

Decision Points:

• Requirements
  • Generate: NCI_Thesaurus_Value_Set.owl file
  • Create convenience method: restrictToMatchingProperties (OR) on multiple matching values.
    • What is the expected input/output?
    • Kim to provide a use case/example for the input/output
  • Load OWL file in less than an hour.
  • Kim to make list of all requirements for generating the OWL file.
• Create a JIRA item to track this issue.