Difference between revisions of "CTS2 RDF Plugin"

From NCBO Wiki
Jump to navigation Jump to search
Line 3: Line 3:
 
=== Known Issues ===
 
=== Known Issues ===
 
* Unbound predicate queries are expensive to execute on the triple store, and CTS2 requires all Properties of a given subject to be displayed. This will cause all CTS2 Read services to have a delay in returning results.
 
* Unbound predicate queries are expensive to execute on the triple store, and CTS2 requires all Properties of a given subject to be displayed. This will cause all CTS2 Read services to have a delay in returning results.
 +
 +
All Read Services are executed by a SPARQL query similar to:
 +
 +
<code>
 +
<nowiki>
 +
<http://some/uri...> ?p ?o
 +
</nowiki>
 +
</code>
 +
 +
This ends up being a very expensive query to the triple store
  
 
* URI Mismatch: Certain portions of the RDF Plugin utilize the NCBO REST service for searching. Sometimes, the URIs returned by the NCBO REST service are not identical to those int the triple store. This will manifest on Entity Read by Name queries.
 
* URI Mismatch: Certain portions of the RDF Plugin utilize the NCBO REST service for searching. Sometimes, the URIs returned by the NCBO REST service are not identical to those int the triple store. This will manifest on Entity Read by Name queries.
Line 15: Line 25:
 
<nowiki>
 
<nowiki>
 
http://purl.bioontology.org/ontology/SNOMED-Ethnic-Group#
 
http://purl.bioontology.org/ontology/SNOMED-Ethnic-Group#
−
</nowiki>
 
−
</code>
 
−
 
−
* Jena Local Name Parsing: Jena is used in the background for parsing the returned RDF.
 
−
 
−
* All Read Services are executed by a SPARQL query similar to:
 
−
 
−
<code>
 
−
<nowiki>
 
−
<http://some/uri...> ?p ?o
 
 
</nowiki>
 
</nowiki>
 
</code>
 
</code>

Revision as of 11:42, 8 June 2012

CTS2 REST Services

Known Issues

  • Unbound predicate queries are expensive to execute on the triple store, and CTS2 requires all Properties of a given subject to be displayed. This will cause all CTS2 Read services to have a delay in returning results.

All Read Services are executed by a SPARQL query similar to:

<http://some/uri...> ?p ?o

This ends up being a very expensive query to the triple store

  • URI Mismatch: Certain portions of the RDF Plugin utilize the NCBO REST service for searching. Sometimes, the URIs returned by the NCBO REST service are not identical to those int the triple store. This will manifest on Entity Read by Name queries.
  • URIs for Entities in Views: Some Entities in Views have different URIs then the resource in the actual code system. For instance

http://informatics.mayo.edu/cts2/services/bioportal-rdf//valueset/SNOMED-ETHNIC-GROUP/definition/SNOMED-ETHNIC-GROUP-43057/resolution/43057

refers to a SNOMEDCT Entity, but Entities have URIs starting with:

http://purl.bioontology.org/ontology/SNOMED-Ethnic-Group#

CodeSystemCatalogQuery

CodeSystemCatalogRead

CodeSystemCatalogVersionQuery

CodeSystemVersionCatalogRead

EntityDescriptionQuery

EntityDescriptionRead

ValueSetCatalogQuery

ValueSetCatalogRead


ResolvedValueSetQuery

AssociationQuery

Retrieved from "https://www.bioontology.org//mediawiki/index.php?title=CTS2_RDF_Plugin&oldid=12301"