edu.stanford.nlp.sequences

Class SequenceSampler

java.lang.Object

edu.stanford.nlp.sequences.SequenceSampler

All Implemented Interfaces:

BestSequenceFinder

public class SequenceSampler
extends Object
implements BestSequenceFinder

This class will sample an output from a sequence model. It assumes that the scores are (unnormalized) log-probabilities. It works by sampling each variable in order, conditioned on the previous variables.

Author:

Jenny Finkel

Constructor Summary

Constructors

Constructor and Description
SequenceSampler()

Method Summary

Methods

Modifier and Type	Method and Description
int[]	bestSequence(SequenceModel ts) Samples each label in turn from left to right.

Methods inherited from class java.lang.Object

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

Constructor Detail

SequenceSampler

public SequenceSampler()

Method Detail

bestSequence

public int[] bestSequence(SequenceModel ts)

Samples each label in turn from left to right.

Specified by:

bestSequence in interface BestSequenceFinder

Returns:

an array containing the int tags of the best sequence