Parsing Engine

danbikel.parser.ms
Class BrokenLeftSubcatModelStructure

java.lang.Object
  extended by danbikel.parser.ProbabilityStructure
      extended by danbikel.parser.ms.SubcatModelStructure1
          extended by danbikel.parser.ms.BrokenLeftSubcatModelStructure
All Implemented Interfaces:
Serializable

public class BrokenLeftSubcatModelStructure
extends SubcatModelStructure1

Provides the complete back-off structure of the subcat-generation model for the left side of the head child. This model structure is just like LeftSubcatModelStructure1 but is “broken” in that its lambdaFudge(int) method returns 5.0 for all back-off levels and its lambdaFudgeTerm(int) method returns 0.0 for all back-off levels, just as Collins had implemented for his thesis parser.

For the actual details of this model's back-off structure, please see SubcatModelStructure1.

See Also:
Serialized Form

Field Summary
 
Fields inherited from class danbikel.parser.ProbabilityStructure
additionalData, defaultModelClassName, defaultModelConstructor, doPruning, estimates, futureList, futures, futuresWithSubcats, histories, historiesWithSubcats, historyList, lambdas, prevHistCount, topLevelCacheSize, transitions
 
Constructor Summary
BrokenLeftSubcatModelStructure()
          Constructs a new BrokenLeftSubcatModelStructure instance.
 
Method Summary
 ProbabilityStructure copy()
          Returns a copy of this object.
 Event getFuture(TrainerEvent trainerEvent, int backOffLevel)
          Gets the future being predicted conditioning on this subcat event.
 double lambdaFudge(int backOffLevel)
          Returns 5.0 regardless of the value of the argument.
 double lambdaFudgeTerm(int backOffLevel)
          Returns 0.0 regardless of the value of the argument.
 
Methods inherited from class danbikel.parser.ms.SubcatModelStructure1
getHistory, maxEventComponents, numLevels
 
Methods inherited from class danbikel.parser.ProbabilityStructure
cacheSize, defaultSmoothingParamsFilename, doCleanup, dontAddNewParameters, doPruning, getAdditionalData, getTopLevelCacheSize, getTransition, jointModel, lambdaPenalty, newModel, priorLevel, removeFuture, removeHistory, removeTransition, saveSmoothingParameters, setAdditionalData, smoothingParametersFile, useSmoothingParameters
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

BrokenLeftSubcatModelStructure

public BrokenLeftSubcatModelStructure()
Constructs a new BrokenLeftSubcatModelStructure instance.

Method Detail

lambdaFudge

public double lambdaFudge(int backOffLevel)
Returns 5.0 regardless of the value of the argument.

Overrides:
lambdaFudge in class SubcatModelStructure1
Parameters:
backOffLevel - the back-off level for which to return a "fudge factor"

lambdaFudgeTerm

public double lambdaFudgeTerm(int backOffLevel)
Returns 0.0 regardless of the value of the argument.

Overrides:
lambdaFudgeTerm in class SubcatModelStructure1

getFuture

public Event getFuture(TrainerEvent trainerEvent,
                       int backOffLevel)
Gets the future being predicted conditioning on this subcat event.

Specified by:
getFuture in class ProbabilityStructure
Parameters:
trainerEvent - the maximal-context event from which to get the future being predicted
backOffLevel - the back-off level whose estimate is being sought
Returns:
the future being predicted conditioning on the left subcat event contained in the speciified TrainerEvent instance

copy

public ProbabilityStructure copy()
Returns a copy of this object.

Specified by:
copy in class ProbabilityStructure

Parsing Engine

Author: Dan Bikel.