Package cc.mallet.fst.confidence
Class MaxEntConfidenceEstimator
- java.lang.Object
-
- cc.mallet.fst.confidence.TransducerConfidenceEstimator
-
- cc.mallet.fst.confidence.MaxEntConfidenceEstimator
-
- All Implemented Interfaces:
java.io.Serializable
public class MaxEntConfidenceEstimator extends TransducerConfidenceEstimator
Estimates the confidence of aSegment
extracted by aTransducer
using aMaxEnt
classifier to classify segments as "correct" or "incorrect." xxx needs some interface work- See Also:
- Serialized Form
-
-
Field Summary
-
Fields inherited from class cc.mallet.fst.confidence.TransducerConfidenceEstimator
model
-
-
Constructor Summary
Constructors Constructor Description MaxEntConfidenceEstimator(Transducer model)
MaxEntConfidenceEstimator(Transducer model, double gaussianVariance)
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description double
estimateConfidenceFor(Segment segment, SumLatticeDefault cachedLattice)
Calculates the confidence in the tagging of aSegment
.MaxEnt
trainClassifier(InstanceList ilist, java.lang.String correct, java.lang.String incorrect)
-
Methods inherited from class cc.mallet.fst.confidence.TransducerConfidenceEstimator
estimateConfidenceFor, getSegmentConfidences, getTransducer, rankSegmentsByConfidence, rankSegmentsByConfidence
-
-
-
-
Constructor Detail
-
MaxEntConfidenceEstimator
public MaxEntConfidenceEstimator(Transducer model, double gaussianVariance)
-
MaxEntConfidenceEstimator
public MaxEntConfidenceEstimator(Transducer model)
-
-
Method Detail
-
trainClassifier
public MaxEnt trainClassifier(InstanceList ilist, java.lang.String correct, java.lang.String incorrect)
-
estimateConfidenceFor
public double estimateConfidenceFor(Segment segment, SumLatticeDefault cachedLattice)
Calculates the confidence in the tagging of aSegment
.- Specified by:
estimateConfidenceFor
in classTransducerConfidenceEstimator
-
-