cc.mallet.fst
Class MultiSegmentationEvaluator

java.lang.Object
  cc.mallet.fst.TransducerEvaluator
      cc.mallet.fst.MultiSegmentationEvaluator

public class MultiSegmentationEvaluator
extends TransducerEvaluator

Evaluates a transducer model, computes the precision, recall and F1 scores; considers segments that span across multiple tokens.

Field Summary

Fields inherited from class cc.mallet.fst.TransducerEvaluator

instanceListDescriptions, instanceLists

Constructor Summary

MultiSegmentationEvaluator(InstanceList[] instanceLists, java.lang.String[] instanceListDescriptions, java.lang.Object[] segmentStartTags, java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator(InstanceList instanceList1, java.lang.String description1, InstanceList instanceList2, java.lang.String description2, InstanceList instanceList3, java.lang.String description3, java.lang.Object[] segmentStartTags, java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator(InstanceList instanceList1, java.lang.String description1, InstanceList instanceList2, java.lang.String description2, java.lang.Object[] segmentStartTags, java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator(InstanceList instanceList1, java.lang.String description1, java.lang.Object[] segmentStartTags, java.lang.Object[] segmentContinueTags)

Method Summary

void	batchTest(InstanceList data, java.util.List<Sequence> predictedSequences, java.lang.String description, java.io.PrintStream viterbiOutputStream) Tests segmentation using an ArrayList of predicted Sequences instead of a Transducer.
void	evaluateInstanceList(TransducerTrainer tt, InstanceList data, java.lang.String description)
int	numIncorrectSegments(Sequence trueOutput, Sequence predOutput) Returns the number of incorrect segments in predOutput

Methods inherited from class cc.mallet.fst.TransducerEvaluator

evaluate, preamble, precondition

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

MultiSegmentationEvaluator

public MultiSegmentationEvaluator(InstanceList[] instanceLists,
                                  java.lang.String[] instanceListDescriptions,
                                  java.lang.Object[] segmentStartTags,
                                  java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator

public MultiSegmentationEvaluator(InstanceList instanceList1,
                                  java.lang.String description1,
                                  java.lang.Object[] segmentStartTags,
                                  java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator

public MultiSegmentationEvaluator(InstanceList instanceList1,
                                  java.lang.String description1,
                                  InstanceList instanceList2,
                                  java.lang.String description2,
                                  java.lang.Object[] segmentStartTags,
                                  java.lang.Object[] segmentContinueTags)

MultiSegmentationEvaluator

public MultiSegmentationEvaluator(InstanceList instanceList1,
                                  java.lang.String description1,
                                  InstanceList instanceList2,
                                  java.lang.String description2,
                                  InstanceList instanceList3,
                                  java.lang.String description3,
                                  java.lang.Object[] segmentStartTags,
                                  java.lang.Object[] segmentContinueTags)

Method Detail

evaluateInstanceList

public void evaluateInstanceList(TransducerTrainer tt,
                                 InstanceList data,
                                 java.lang.String description)

Specified by:

evaluateInstanceList in class TransducerEvaluator

numIncorrectSegments

public int numIncorrectSegments(Sequence trueOutput,
                                Sequence predOutput)

Returns the number of incorrect segments in predOutput

Parameters:

trueOutput - truth

predOutput - predicted

Returns:

number of incorrect segments

batchTest

public void batchTest(InstanceList data,
                      java.util.List<Sequence> predictedSequences,
                      java.lang.String description,
                      java.io.PrintStream viterbiOutputStream)

Tests segmentation using an ArrayList of predicted Sequences instead of a Transducer. If predictedSequence is null, don't include in stats (useful for error analysis).

Parameters:

data - list of instances to be segmented

predictedSequences - predictions

description - description of trial

viterbiOutputStream - where to print the Viterbi paths