|
Parsing Engine | |||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectdanbikel.parser.Model
danbikel.parser.InterpolatedKnesserNeyModel
public class InterpolatedKnesserNeyModel
Implements a model that uses interpolated Knesser-Ney smoothing.
Field Summary | |
---|---|
protected double |
optimalDiscountEstimate
|
Constructor Summary | |
---|---|
InterpolatedKnesserNeyModel(ProbabilityStructure structure)
Constructs a Model instance that uses interpolated Knesser-Ney
smoothing instead of the default smoothing method when estimating
probabilities. |
Method Summary | |
---|---|
void |
deriveCounts(CountsTable trainerCounts,
Filter filter,
double threshold,
FlexibleMap canonical,
boolean deriveOtherModelCounts)
Derives all counts from the specified counts table, using the probability structure specified in the constructor. |
protected double |
estimateProb(ProbabilityStructure probStructure,
TrainerEvent event)
Returns the smoothed probability estimate of a transition contained in the specified TrainerEvent object. |
protected void |
precomputeProbs(MapToPrimitive.Entry transEntry,
double[] lambdas,
double[] estimates,
Transition[] transitions,
Event[] histories,
int lastLevel)
Precomputes the probabilities and smoothing values for the Transition object contained as a key within the specified
map entry, where the value is the count of the transition. |
protected void |
precomputeProbs(TrainerEvent event,
Transition[] transitions,
Event[] histories)
Deprecated. This method is called by Model.precomputeProbs(CountsTable,Filter) , which is also deprecated. |
protected void |
storePrecomputedProbs(double[] lambdas,
double[] estimates,
Transition[] transitions,
Event[] histories,
int lastLevel)
Stores the specified smoothing values (lambdas) and smoothed probability estimates in the Model.precomputedProbs and Model.smoothingParams
map arrays. |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Field Detail |
---|
protected double optimalDiscountEstimate
Constructor Detail |
---|
public InterpolatedKnesserNeyModel(ProbabilityStructure structure)
Model
instance that uses interpolated Knesser-Ney
smoothing instead of the default smoothing method when estimating
probabilities.
structure
- the probability structure for which this model will
estimate probabilitiesMethod Detail |
---|
public void deriveCounts(CountsTable trainerCounts, Filter filter, double threshold, FlexibleMap canonical, boolean deriveOtherModelCounts)
Model
deriveCounts
in class Model
trainerCounts
- a map from TrainerEvent
objects to
their counts (as double
s) from which to derive countsfilter
- used to filter out TrainerEvent
objects
whose derived counts should not be derived for this modelthreshold
- a (currently unused) count cut-off thresholdcanonical
- a reflexive map used to canonicalize objects
created when deriving countsderiveOtherModelCounts
- an unused parameter, as this class
does not contain other, internal Model
instancesprotected double estimateProb(ProbabilityStructure probStructure, TrainerEvent event)
TrainerEvent
object. The smoothing method employed
will be the interpolated version of Knesser-Ney.
estimateProb
in class Model
probStructure
- a ProbabilityStructure
object that is
either Model.structure
or a copy of it, used for
temporary storage during the computation performed by
this methodevent
- the TrainerEvent
containing a transition
from a history to a future whose smoothed probability
is to be computed
TrainerEvent
objectprotected void precomputeProbs(MapToPrimitive.Entry transEntry, double[] lambdas, double[] estimates, Transition[] transitions, Event[] histories, int lastLevel)
Model
Transition
object contained as a key within the specified
map entry, where the value is the count of the transition.
precomputeProbs
in class Model
transEntry
- a map entry mapping a Transition
object to its count (a double
)lambdas
- an array in which to store the smoothing value for
each of the back-off levelsestimates
- an array in which to store the maximum-likelihood
estimate at each of the back-off levelstransitions
- an array in which to store the Transition
instance for each of the back-off levelshistories
- an array in which to store the history, an
Event
instance, for each of the back-off levelslastLevel
- the last back-off level (the value equal to
Model.numLevels
- 1
)Model.precomputeProbs()
protected void storePrecomputedProbs(double[] lambdas, double[] estimates, Transition[] transitions, Event[] histories, int lastLevel)
Model
Model.precomputedProbs
and Model.smoothingParams
map arrays.
storePrecomputedProbs
in class Model
lambdas
- an array containing the smoothing value for each of the
back-off levelsestimates
- an array containing the maximum-likelihood estimate at
each of the back-off levelstransitions
- an array containing the Transition
instance for each of the back-off levelshistories
- an array in which to store the history, an
Event
instance, for each of the back-off levelslastLevel
- the last back-off level (the value equal to
Model.numLevels
- 1
)Model.precomputeProbs()
protected void precomputeProbs(TrainerEvent event, Transition[] transitions, Event[] histories)
Model.precomputeProbs(CountsTable,Filter)
, which is also deprecated.
precomputeProbs
in class Model
event
- the TrainerEvent
object from which probabilities
are to be precomputedtransitions
- temporary storage to be used during an invocation
of this methodhistories
- temporary storage to be used during an invocation
of this method
|
Parsing Engine | |||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |