Package com.basistech.rosette.dm

Class CategorizerResult

java.lang.Object

com.basistech.rosette.dm.BaseAttribute

com.basistech.rosette.dm.CategorizerResult

All Implemented Interfaces:

Serializable

public class CategorizerResult
extends BaseAttribute
implements Serializable

Associates a label with a document. Various applications may generate these labels. For an article about Michael Jordan, a categorzier might produce a label like "SPORTS", and sentiment analyzer might produce "neg", "neu", "pos", while a topic generator might produce "basketball", "bulls", "chicago". Applications typically produce a ranked list of CategorizerResults. Each result has a label. Optional fields include confidence, an explanation set, and per-feature scores. This result is currently used for document categorization, sentiment analysis, and topic generation.

See Also:

• Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	CategorizerResult.Builder	A builder for classifier results.

Field Summary

Fields inherited from class com.basistech.rosette.dm.BaseAttribute

extendedProperties

Method Summary

Modifier and Type	Method	Description
Double	getConfidence()	Returns the confidence this result.
List<String>	getExplanationSet()	Returns a list of the N highest positive features that contributed to the categorization (the 'explanation set'), ordered by descending score.
String	getLabel()	Returns the label of this result.
Map<String,Double>	getPerFeatureScores()	Returns a map of feature string to raw score for this result.
Double	getScore()	Returns the raw score of this result.
protected com.google.common.base.MoreObjects.ToStringHelper	toStringHelper()

Methods inherited from class com.basistech.rosette.dm.BaseAttribute

getExtendedProperties, listOrNull, setExtendedProperty, toString

Methods inherited from class java.lang.Object

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

Method Details

getLabel

public String getLabel()

Returns the label of this result.

Returns:

the label of this result

getScore

public Double getScore()

Returns the raw score of this result. The meaning of this value is not documented. Typical users should use getConfidence() instead.

Returns:

the raw score of this result

getConfidence

public Double getConfidence()

Returns the confidence this result. If confidence calculations were enabled, the value returned will be in the range [0.0, 1.0]. Otherwise null will be returned.

Returns:

the confidence of this result

getExplanationSet

public List<String> getExplanationSet()

Returns a list of the N highest positive features that contributed to the categorization (the 'explanation set'), ordered by descending score. Returns null explanations are disabled. The value of N is determined by the calling Categorizer.

Returns:

the explanation set

getPerFeatureScores

public Map<String,Double> getPerFeatureScores()

Returns a map of feature string to raw score for this result. The meaning of a score is not documented. This is an expert-level feature used for debugging and error analysis.

Returns:

the per-feature scores

toStringHelper

protected com.google.common.base.MoreObjects.ToStringHelper toStringHelper()

Overrides:

toStringHelper in class BaseAttribute