edu.stanford.nlp.parser.shiftreduce

Class PerceptronModel

java.lang.Object

edu.stanford.nlp.parser.shiftreduce.BaseModel

edu.stanford.nlp.parser.shiftreduce.PerceptronModel

All Implemented Interfaces:

Serializable

public class PerceptronModel
extends BaseModel

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
PerceptronModel(PerceptronModel other)
PerceptronModel(ShiftReduceOptions op, Index<Transition> transitionIndex, Set<String> knownStates, Set<String> rootStates, Set<String> rootOnlyStates)

Method Summary

Modifier and Type	Method and Description
void	averageModels(Collection<PerceptronModel> models)
void	averageScoredModels(Collection<ScoredObject<PerceptronModel>> scoredModels)
Collection<ScoredObject<Integer>>	findHighestScoringTransitions(State state, boolean requireLegal, int numTransitions, List<ParserConstraint> constraints)
void	outputStats() Output some random facts about the model
void	trainModel(String serializedPath, Tagger tagger, Random random, List<Tree> binarizedTrees, List<List<Transition>> transitionLists, Treebank devTreebank, int nThreads) Will train the model on the given treebank, using devTreebank as a dev set.

Methods inherited from class edu.stanford.nlp.parser.shiftreduce.BaseModel

findEmergencyTransition

Methods inherited from class java.lang.Object

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

Constructor Detail

PerceptronModel

public PerceptronModel(ShiftReduceOptions op,
                       Index<Transition> transitionIndex,
                       Set<String> knownStates,
                       Set<String> rootStates,
                       Set<String> rootOnlyStates)

PerceptronModel

public PerceptronModel(PerceptronModel other)

Method Detail

averageScoredModels

public void averageScoredModels(Collection<ScoredObject<PerceptronModel>> scoredModels)

averageModels

public void averageModels(Collection<PerceptronModel> models)

outputStats

public void outputStats()

Output some random facts about the model

findHighestScoringTransitions

public Collection<ScoredObject<Integer>> findHighestScoringTransitions(State state,
                                                                       boolean requireLegal,
                                                                       int numTransitions,
                                                                       List<ParserConstraint> constraints)

Specified by:

findHighestScoringTransitions in class BaseModel

trainModel

public void trainModel(String serializedPath,
                       Tagger tagger,
                       Random random,
                       List<Tree> binarizedTrees,
                       List<List<Transition>> transitionLists,
                       Treebank devTreebank,
                       int nThreads)

Will train the model on the given treebank, using devTreebank as a dev set. If op.retrainAfterCutoff is set, will rerun training after the first time through on a limited set of features.

Specified by:

trainModel in class BaseModel

Parameters:

serializedPath - Where serialized models go. If the appropriate options are set, the method can use this to save intermediate models.

tagger - The tagger to use when evaluating devTreebank. TODO: it would make more sense for ShiftReduceParser to retag the trees first

random - A random number generator to use for any random numbers. Useful to make sure results can be reproduced.

binarizedTrees - The treebank to train from.

transitionLists - binarizedTrainTrees converted into lists of transitions that will reproduce the same tree.

devTreebank - a set of trees which can be used for dev testing (assuming the user provided a dev treebank)

nThreads - how many threads the model can use for training