edu.stanford.nlp.ling
Class CoreAnnotations.StackedNamedEntityTagAnnotation

java.lang.Object
  extended by edu.stanford.nlp.ling.CoreAnnotations.StackedNamedEntityTagAnnotation
All Implemented Interfaces:
CoreAnnotation<java.lang.String>, TypesafeMap.Key<CoreMap,java.lang.String>
Enclosing class:
CoreAnnotations

public static class CoreAnnotations.StackedNamedEntityTagAnnotation
extends java.lang.Object
implements CoreAnnotation<java.lang.String>

The CoreMap key for getting the token-level named entity tag (e.g., DATE, PERSON, etc.) from a previous NER tagger. NERFeatureFactory is sensitive to this tag and will turn the annotations from the previous NER tagger into new features. This is currently used to implement one level of stacking -- we may later change it to take a list as needed. This key is typically set on token annotations.


Constructor Summary
CoreAnnotations.StackedNamedEntityTagAnnotation()
           
 
Method Summary
 java.lang.Class<java.lang.String> getType()
          Returns the type associated with this annotation.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

CoreAnnotations.StackedNamedEntityTagAnnotation

public CoreAnnotations.StackedNamedEntityTagAnnotation()
Method Detail

getType

public java.lang.Class<java.lang.String> getType()
Description copied from interface: CoreAnnotation
Returns the type associated with this annotation. This method must return the same class type as its value type parameter. It feels like one should be able to get away without this method, but because Java erases the generic type signature, that info disappears at runtime.

Specified by:
getType in interface CoreAnnotation<java.lang.String>


Stanford NLP Group