Package opennlp.maxent

Provides main functionality of the maxent package including data structures and algorithms for parameter estimation.

See:
          Description

Interface Summary
ContextGenerator Generate contexts for maxent decisions.
DataIndexer Object which compresses events in memory and performs feature selection.
DataStream A interface for objects which can deliver a stream of training data to be supplied to an EventStream.
Evalable Interface for components which use maximum entropy models and can evaluate the performace of the models using the TrainEval class.
EventCollector An interface for objects which read events during training.
EventStream A object which can deliver a stream of training events for the GIS procedure (or others such as IIS if and when they are implemented).
MaxentModel Interface for maximum entropy models.
ModelDomain A simple interface that represents a domain to which a particular maxent model is primarily applicable.
ModelSetter A object to facilitate the resetting of a MaxentModel variable to a new value (model).
Prior This interface allows one to implement a prior distribution for use in maximum entropy model training.
 

Class Summary
AbstractDataIndexer Abstract class for collecting event and context counts used in training.
BasicContextGenerator Generate contexts for maxent decisions, assuming that the input given to the getContext() method is a String containing contextual predicates separated by spaces.
BasicEventStream A object which can deliver a stream of training events assuming that each event is represented as a space separated list containing all the contextual predicates, with the last item being the outcome.
BinToAscii A program to convert from java binary doubles to ascii
ComparableEvent A maxent event representation which we can use to sort based on the predicates indexes contained in the events.
ComparablePredicate A maxent predicate representation which we can use to sort based on the outcomes.
Context Class which associates a real valueed parameter or expected value with a particular contextual predicate or feature.
Counter A simple class which is essentially an Integer which is mutable via incrementation.
DomainToModelMap A class which stores a mapping from ModelDomain objects to MaxentModels.
EvalParameters This class encapsulates the varibales used in producing probabilities from a model and facilitaes passing these variables to the eval method.
Event The context of a decision point during training.
EventCollectorAsStream A wrapper to turn EventCollectors created for Maxent 1.0 into EventStreams for Maxent 1.2.
FileEventStream Class for using a file of events as an event stream.
GIS A Factory class which uses instances of GISTrainer to create and train GISModels.
GISModel A maximum entropy model which has been trained using the Generalized Iterative Scaling procedure (implemented in GIS.java).
IntegerPool A pool of read-only, unsigned Integer objects within a fixed, non-sparse range.
Main Main file for opennlp.maxent.
ModelReplacementManager A object which can be used to ensure that a Maxent application can swap the model currently in use with a new one in a thread-safe manner without stopping the servicing of requests.
MutableContext Class used to store parameters or expected values associated with this context which can be updated or assigned.
OnePassDataIndexer An indexer for maxent model data which handles cutoffs for uncommon contextual predicates and provides a unique integer index for each of the predicates.
OnePassRealValueDataIndexer An indexer for maxent model data which handles cutoffs for uncommon contextual predicates and provides a unique integer index for each of the predicates and maintains event values.
PlainTextByLineDataStream This DataStream implementation will take care of reading a plain text file and returning the Strings between each new line character, which is what many Maxent applications need in order to create EventStreams.
RealBasicEventStream  
RealValueFileEventStream  
TIntParamHashMap Data structure for storing a models parameters for each outcome associated with a specific context.
TObjectIndexHashMap Data structure storing the mapping between a context and its integer value.
TrainEval Trains or evaluates maxent components which have implemented the Evalable interface.
TwoPassDataIndexer Collecting event and context counts by making two passes over the events.
UniformPrior Provide a maximum entropy model with a uniform prior.
 

Package opennlp.maxent Description

Provides main functionality of the maxent package including data structures and algorithms for parameter estimation.



Copyright © 2005 Jason Baldridge, Gann Bierner, and Thomas Morton. All Rights Reserved.