Package opennlp.tools.util.featuregen
Class BrownCluster
java.lang.Object
opennlp.tools.util.featuregen.BrownCluster
- All Implemented Interfaces:
SerializableArtifact
Class to load a Brown cluster document: word\tword_class\tprob
Originally available at: http://metaoptimize.com/projects/wordreprs/. Further details can be found in the related research paper.
The file containing the clustering lexicon has to be passed as the
value of the dict attribute of each BrownCluster
feature generator.
-
Nested Class Summary
-
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptionClass<?>
Retrieves the class which can serialize and recreate this artifact.lookupToken
(String string) Check if a token is in the Brown:paths, token map.void
serialize
(OutputStream out)
-
Constructor Details
-
BrownCluster
Generates the token to cluster map from Brown cluster anInputStream
.Note: we only add those tokens with frequency bigger than
5
.- Parameters:
in
- A valid, openInputStream
to read from.- Throws:
IOException
- the io exception
-
-
Method Details
-
lookupToken
Check if a token is in the Brown:paths, token map.- Parameters:
string
- the token to look-up- Returns:
- the brown class if such token is in the brown cluster map
-
serialize
- Throws:
IOException
-
getArtifactSerializerClass
Description copied from interface:SerializableArtifact
Retrieves the class which can serialize and recreate this artifact.Note: The serializer class must have a
public zero argument constructor
or an exception is thrown during model serialization/loading.- Specified by:
getArtifactSerializerClass
in interfaceSerializableArtifact
- Returns:
- The corresponding
ArtifactSerializer
class.
-