incubator-ctakes-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kim Ebert <kim.eb...@perfectsearchcorp.com>
Subject Re: Missing conceptType between multiple runs of same data corpus
Date Wed, 23 Jan 2013 04:40:06 GMT
It appears this may be caused by looking at on the first item in the 
array that is passed to the lookupConceptType function in 
org.mitre.medfacts.uima.assertion.ConceptLookup

   public static ConceptType lookupConceptType(FSArray ontologyConceptArray)
   {
     //logger.info("begin ConceptLookup.lookupConceptType");
     FeatureStructure firstConceptFS = null;
     boolean hasConcept =
       (ontologyConceptArray != null &&
        ontologyConceptArray.size() >= 1 &&
*       ontologyConceptArray.get(0) instanceof OntologyConcept*
       );
     if (hasConcept)
     {
*      firstConceptFS = ontologyConceptArray.get(0);*
     }

Is there a reason to look at only the first concept? It appears in 
testing that this array is unsorted? I've done some testing and found 
that sometimes the first item in the array doesn't provide a concept 
type, but the second or third item in the array may provide a concept type.

Kim Ebert
1.801.669.7342
Perfect Search Corp
http://www.perfectsearchcorp.com/


On 01/22/2013 01:00 PM, Kim Ebert wrote:
> Hi All,
>
> I've been doing some testing on cTAKES 2.5 and I have noticed that 
> occasionally the conceptType is dropped from the output of a document 
> while using the AggregatePlaintextUMLSProcessor.
>
> <org.mitre.medfacts.type.Concept ... conceptType="PROBLEM" ... /> <!-- 
> conceptType is present -->
>
> vs
>
> <org.mitre.medfacts.type.Concept ...  /> <!-- no conceptType -->
>
> I was wondering if anyone else has seen this type of behavior?
>
> Thanks,
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message