Difference between revisions of "ICatLinearization"

From Protege Wiki
Jump to: navigation, search
Line 2: Line 2:
 
== Overview ==
 
== Overview ==
 
* Linearization in iCat allows the same entity to be a child of two parents, but only counted as a child of one parent for particular purposes.  
 
* Linearization in iCat allows the same entity to be a child of two parents, but only counted as a child of one parent for particular purposes.  
* For example, the same code could be used for a particular disease, but have different parents depending on if it is entered for the purpose of morbidity or for mortality.
+
* For example, Bacterial Pneumonia is both an infection and a respiratory disease. The same code can be used to describe this disease, but have different parents depending on whether the same code could be used for a particular disease, but have different parents depending on if it is entered for the purpose of morbidity or for mortality.
* This directly affects the way in which statistics are aggregated under particular categories. If there are different parents for morbidity and mortality, then we might have the same code belong to a different category (even chapter) and any incidents of the code would affect the total number of incidents attributed to that category for the pur
+
* This directly affects the way in which statistics are aggregated under particular categories. In some cases (say a classification of morbidity) we might want to classify Bacterial Pneumonia as a disease. In that case, all incidents of Bacterial Pneumonia add one count to the tally of all diseases. If we were more interested in infections, then each case of Bacterial Pneumonia would be counted as an infection.

Revision as of 20:58, July 14, 2010

iCat Linearization

Overview

  • Linearization in iCat allows the same entity to be a child of two parents, but only counted as a child of one parent for particular purposes.
  • For example, Bacterial Pneumonia is both an infection and a respiratory disease. The same code can be used to describe this disease, but have different parents depending on whether the same code could be used for a particular disease, but have different parents depending on if it is entered for the purpose of morbidity or for mortality.
  • This directly affects the way in which statistics are aggregated under particular categories. In some cases (say a classification of morbidity) we might want to classify Bacterial Pneumonia as a disease. In that case, all incidents of Bacterial Pneumonia add one count to the tally of all diseases. If we were more interested in infections, then each case of Bacterial Pneumonia would be counted as an infection.