Package dragon.ml.seqmodel.feature

Various feature types for conditional random field applications.

See:
Description

Interface Summary

Feature	A single feature returned by the FeatureGenerator needs to support this interface.
FeatureDictionary	Interface of Feature Dictionary
FeatureGenerator	Interface of Feature Generator
FeatureType	Interface of feature type which is actually a factory generating certain type of features

Class Summary

AbstractFeatureDictionary	Abstract class for feature dictionary
AbstractFeatureType	Abstract class for feature type
AbstractFeatureTypeWrapper	Abstract class for feature type wrapper
BasicFeature	Basic data structure for feature data
BasicFeatureGenerator	The FeatureGenerator is an aggregator over all these different feature types.
FeatureDictionaryChar	Character based feature dictionary
FeatureDictionaryNum	Numarical feature dictionary
FeatureIdentifier	Feature identifier
FeatureMap	Feature map
FeatureTypeConcatRegex	ConcatRegexFeatures generates features by matching the token with the character patterns.
FeatureTypeEdge	Feature type edge
FeatureTypeEnd	Feature type for end
FeatureTypeKnownWord	Feature type of known word
FeatureTypePosition	This can be used as a wrapper around a FeatureType class that wants to generate features which take into account the normalized distance of the token to the start of a segement.
FeatureTypePOSPattern	This feature type will create features about part of speech patterns for segments This feature type should be wrapped by FeatureTypeStateLoop.
FeatureTypePrior	This feature type prior
FeatureTypeRegexCount	This FeatureType can not be used independently because the label is not set.
FeatureTypeSegmentLength	This FeatureType can not be used independently because the label is not set.
FeatureTypeSegmentLengthPoly	This FeatureType can not be used independently because the label is not set.
FeatureTypeStart	The feature type for start
FeatureTypeStateLoop	This can be used as a wrapper around a FeatureType class that wants to generate a feature for each state.
FeatureTypeTokenLoop	This can be used as a wrapper around a FeatureType class that wants to generate features for each token in a segment.
FeatureTypeUnknownWord	Feature type for unknown world
FeatureTypeWindow	Window feature type
FeatureTypeWord	World feature type
FeatureTypeWordScore	These return one feature per state.
Window	Data structor for a window

Package dragon.ml.seqmodel.feature Description

Various feature types for conditional random field applications.

Package Specification

A feature type is actually a feature generator. It can generate a given type of features from a fragment of a sequence. One should implement the interface called FeatureType to create own feature types. For saving time, one can extend the Abstract FeatureType instead of coding from the sctratch.