com.exalead.mercury.mami.learning.v10

Class ClassifierConfig

java.lang.Object

com.exalead.mercury.mami.learning.v10.UnSupervisedLearningConfig

com.exalead.mercury.mami.learning.v10.SupervisedLearningConfig

com.exalead.mercury.mami.learning.v10.ClassifierConfig

All Implemented Interfaces:

com.exalead.util.Checkable, java.io.Serializable

public class ClassifierConfig
extends SupervisedLearningConfig
implements com.exalead.util.Checkable, java.io.Serializable

Specify a classification job to be performed in this learning job.
A Classification job extends automatically a pre-existing classification It takes as input an already-classified document having a meta specifing its category. (the name of the meta to consider is specified in the configuration), and learns from these samples documents how to classify.
New documents are processed and classified against the existing configuration.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	ClassifierConfig.Type

Nested classes/interfaces inherited from class com.exalead.mercury.mami.learning.v10.SupervisedLearningConfig

SupervisedLearningConfig.Categories

Nested classes/interfaces inherited from class com.exalead.mercury.mami.learning.v10.UnSupervisedLearningConfig

UnSupervisedLearningConfig.Conditions, UnSupervisedLearningConfig.Descriptors, UnSupervisedLearningConfig.SkippedAnnotations

Field Summary

Fields

Modifier and Type	Field and Description
static int	DEFAULT_MAX_CATEGORIES
static int	DEFAULT_MIN_FREQ
static int	DEFAULT_MIN_FREQ_IN_CLASS
static int	DEFAULT_MIN_KEYWORDS
static double	DEFAULT_MIN_SCORE
int	maxCategories
int	minFreq
int	minFreqInClass
int	minKeywords
double	minScore
java.lang.String	type

Fields inherited from class com.exalead.mercury.mami.learning.v10.SupervisedLearningConfig

categories, meta

Fields inherited from class com.exalead.mercury.mami.learning.v10.UnSupervisedLearningConfig

conditions, DEFAULT_MAX_DOCUMENT_THRESHOLD, DEFAULT_MIN_DOCUMENT_REQUIRED, descriptors, instance, managedResourceGroup, maxDocumentThreshold, minDocumentRequired, skippedAnnotations

Constructor Summary

Constructors

Constructor and Description
ClassifierConfig()
ClassifierConfig(ClassifierConfig o) Copy constructor

Method Summary

Modifier and Type	Method and Description
void	check(boolean deep, java.lang.String errorContext) Checks this ClassifierConfig.
static ClassifierConfig	fromString(java.lang.String s) String representation of this ClassifierConfig.
int	getMaxCategories() The maximum number of categories to set during prediction. The classification module can output several classifications a given document.
int	getMinFreq() The minimum global frequency for a keyword
int	getMinFreqInClass() The minimum frequency for a keyword in a class
int	getMinKeywords() The minimum number of keywords required to launch the learning on the document. A keyword can be a word or an annotation tag depending on the attribute @matchedAnnotations. Each keyword will be processed by the learning processor.
double	getMinScore() The minimum score for a keyword
java.lang.String	getType() The type of classifier (bayesian or hbayesian) Value can be one of bayesian hbayesian
ClassifierConfig.Type	getTypeAsEnum() The type of classifier (bayesian or hbayesian) Value can be one of bayesian hbayesian
static boolean	isA(java.lang.String str, ClassifierConfig.Type item)
static boolean	isOneOf(java.lang.String str, ClassifierConfig.Type... items)
ClassifierConfig	makeCopy() Creates and returns a deep copy of this ClassifierConfig.
static ClassifierConfig	readFrom(java.io.InputStream is) Read this ClassifierConfig from an XML fragment.
void	setMaxCategories(int maxCategories) The maximum number of categories to set during prediction. The classification module can output several classifications a given document.
void	setMinFreq(int minFreq) The minimum global frequency for a keyword
void	setMinFreqInClass(int minFreqInClass) The minimum frequency for a keyword in a class
void	setMinKeywords(int minKeywords) The minimum number of keywords required to launch the learning on the document. A keyword can be a word or an annotation tag depending on the attribute @matchedAnnotations. Each keyword will be processed by the learning processor.
void	setMinScore(double minScore) The minimum score for a keyword
void	setType(ClassifierConfig.Type item) The type of classifier (bayesian or hbayesian) Value can be one of bayesian hbayesian
void	setType(java.lang.String type) The type of classifier (bayesian or hbayesian) Value can be one of bayesian hbayesian
java.lang.String	toString() String representation of this ClassifierConfig.
ClassifierConfig	withCategories(Category... __values) The list of categories.
ClassifierConfig	withCategories(java.util.Collection<Category> __values)
ClassifierConfig	withConditions(java.util.Collection<LearningTriggerCondition> __values)
ClassifierConfig	withConditions(LearningTriggerCondition... __values) The list of triggering conditions. This learning job is triggered as soon as any of this condition is fulfilled.
ClassifierConfig	withDescriptors(java.util.Collection<Descriptor> __values)
ClassifierConfig	withDescriptors(Descriptor... __values) The list of MOT annotation used to extract descriptors (if empty, words are used).
ClassifierConfig	withInstance(java.lang.String instance)
ClassifierConfig	withManagedResourceGroup(java.lang.String managedResourceGroup)
ClassifierConfig	withMaxCategories(int maxCategories)
ClassifierConfig	withMaxCategories(java.lang.Integer maxCategories)
ClassifierConfig	withMaxDocumentThreshold(int maxDocumentThreshold)
ClassifierConfig	withMaxDocumentThreshold(java.lang.Integer maxDocumentThreshold)
ClassifierConfig	withMeta(java.lang.String meta)
ClassifierConfig	withMinDocumentRequired(int minDocumentRequired)
ClassifierConfig	withMinDocumentRequired(java.lang.Integer minDocumentRequired)
ClassifierConfig	withMinFreq(int minFreq)
ClassifierConfig	withMinFreq(java.lang.Integer minFreq)
ClassifierConfig	withMinFreqInClass(int minFreqInClass)
ClassifierConfig	withMinFreqInClass(java.lang.Integer minFreqInClass)
ClassifierConfig	withMinKeywords(int minKeywords)
ClassifierConfig	withMinKeywords(java.lang.Integer minKeywords)
ClassifierConfig	withMinScore(double minScore)
ClassifierConfig	withSkippedAnnotations(java.util.Collection<StringValue> __values)
ClassifierConfig	withSkippedAnnotations(StringValue... __values) The list of MOT annotation used to ignore tokens.
ClassifierConfig	withType(java.lang.String type)
void	writeTo(java.io.OutputStream os) Write this ClassifierConfig as an XML fragment

Methods inherited from class com.exalead.mercury.mami.learning.v10.SupervisedLearningConfig

getCategories, getMeta, setCategories, setMeta, withCategories

Methods inherited from class com.exalead.mercury.mami.learning.v10.UnSupervisedLearningConfig

getConditions, getDescriptors, getInstance, getManagedResourceGroup, getMaxDocumentThreshold, getMinDocumentRequired, getSkippedAnnotations, setConditions, setDescriptors, setInstance, setManagedResourceGroup, setMaxDocumentThreshold, setMinDocumentRequired, setSkippedAnnotations, withConditions, withDescriptors, withSkippedAnnotations

Methods inherited from class java.lang.Object

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

Field Detail

type

public java.lang.String type

minKeywords

public int minKeywords

DEFAULT_MIN_KEYWORDS

public static final int DEFAULT_MIN_KEYWORDS

See Also:

Constant Field Values

maxCategories

public int maxCategories

DEFAULT_MAX_CATEGORIES

public static final int DEFAULT_MAX_CATEGORIES

See Also:

Constant Field Values

minFreq

public int minFreq

DEFAULT_MIN_FREQ

public static final int DEFAULT_MIN_FREQ

See Also:

Constant Field Values

minFreqInClass

public int minFreqInClass

DEFAULT_MIN_FREQ_IN_CLASS

public static final int DEFAULT_MIN_FREQ_IN_CLASS

See Also:

Constant Field Values

minScore

public double minScore

DEFAULT_MIN_SCORE

public static final double DEFAULT_MIN_SCORE

See Also:

Constant Field Values

Constructor Detail

ClassifierConfig

public ClassifierConfig()

ClassifierConfig

public ClassifierConfig(ClassifierConfig o)

Copy constructor

Method Detail

isA

public static boolean isA(java.lang.String str,
                          ClassifierConfig.Type item)

isOneOf

public static boolean isOneOf(java.lang.String str,
                              ClassifierConfig.Type... items)

setType

public void setType(ClassifierConfig.Type item)

The type of classifier (bayesian or hbayesian) Value can be one of

• bayesian
• hbayesian

getTypeAsEnum

public ClassifierConfig.Type getTypeAsEnum()

The type of classifier (bayesian or hbayesian) Value can be one of

• bayesian
• hbayesian

withMeta

public ClassifierConfig withMeta(java.lang.String meta)

Overrides:

withMeta in class SupervisedLearningConfig

withCategories

public ClassifierConfig withCategories(Category... __values)

The list of categories.

Overrides:

withCategories in class SupervisedLearningConfig

withCategories

public ClassifierConfig withCategories(java.util.Collection<Category> __values)

Overrides:

withCategories in class SupervisedLearningConfig

withInstance

public ClassifierConfig withInstance(java.lang.String instance)

Overrides:

withInstance in class SupervisedLearningConfig

withMinDocumentRequired

public ClassifierConfig withMinDocumentRequired(int minDocumentRequired)

Overrides:

withMinDocumentRequired in class SupervisedLearningConfig

withMinDocumentRequired

public ClassifierConfig withMinDocumentRequired(java.lang.Integer minDocumentRequired)

Overrides:

withMinDocumentRequired in class SupervisedLearningConfig

withMaxDocumentThreshold

public ClassifierConfig withMaxDocumentThreshold(int maxDocumentThreshold)

Overrides:

withMaxDocumentThreshold in class SupervisedLearningConfig

withMaxDocumentThreshold

public ClassifierConfig withMaxDocumentThreshold(java.lang.Integer maxDocumentThreshold)

Overrides:

withMaxDocumentThreshold in class SupervisedLearningConfig

withDescriptors

public ClassifierConfig withDescriptors(Descriptor... __values)

The list of MOT annotation used to extract descriptors (if empty, words are used).

Overrides:

withDescriptors in class SupervisedLearningConfig

withDescriptors

public ClassifierConfig withDescriptors(java.util.Collection<Descriptor> __values)

Overrides:

withDescriptors in class SupervisedLearningConfig

withSkippedAnnotations

public ClassifierConfig withSkippedAnnotations(StringValue... __values)

The list of MOT annotation used to ignore tokens.

Overrides:

withSkippedAnnotations in class SupervisedLearningConfig

withSkippedAnnotations

public ClassifierConfig withSkippedAnnotations(java.util.Collection<StringValue> __values)

Overrides:

withSkippedAnnotations in class SupervisedLearningConfig

withConditions

public ClassifierConfig withConditions(LearningTriggerCondition... __values)

The list of triggering conditions.
This learning job is triggered as soon as any of this condition is fulfilled.

Overrides:

withConditions in class SupervisedLearningConfig

withConditions

public ClassifierConfig withConditions(java.util.Collection<LearningTriggerCondition> __values)

Overrides:

withConditions in class SupervisedLearningConfig

withManagedResourceGroup

public ClassifierConfig withManagedResourceGroup(java.lang.String managedResourceGroup)

Overrides:

withManagedResourceGroup in class SupervisedLearningConfig

setType

public void setType(java.lang.String type)

The type of classifier (bayesian or hbayesian) Value can be one of

• bayesian
• hbayesian

getType

public java.lang.String getType()

The type of classifier (bayesian or hbayesian) Value can be one of

• bayesian
• hbayesian

withType

public ClassifierConfig withType(java.lang.String type)

setMinKeywords

public void setMinKeywords(int minKeywords)

The minimum number of keywords required to launch the learning on the document.
A keyword can be a word or an annotation tag depending on the attribute @matchedAnnotations.
Each keyword will be processed by the learning processor.

getMinKeywords

public int getMinKeywords()

The minimum number of keywords required to launch the learning on the document.
A keyword can be a word or an annotation tag depending on the attribute @matchedAnnotations.
Each keyword will be processed by the learning processor.

withMinKeywords

public ClassifierConfig withMinKeywords(int minKeywords)

withMinKeywords

public ClassifierConfig withMinKeywords(java.lang.Integer minKeywords)

setMaxCategories

public void setMaxCategories(int maxCategories)

The maximum number of categories to set during prediction.
The classification module can output several classifications a given document.

getMaxCategories

public int getMaxCategories()

The maximum number of categories to set during prediction.
The classification module can output several classifications a given document.

withMaxCategories

public ClassifierConfig withMaxCategories(int maxCategories)

withMaxCategories

public ClassifierConfig withMaxCategories(java.lang.Integer maxCategories)

setMinFreq

public void setMinFreq(int minFreq)

The minimum global frequency for a keyword

getMinFreq

public int getMinFreq()

The minimum global frequency for a keyword

withMinFreq

public ClassifierConfig withMinFreq(int minFreq)

withMinFreq

public ClassifierConfig withMinFreq(java.lang.Integer minFreq)

setMinFreqInClass

public void setMinFreqInClass(int minFreqInClass)

The minimum frequency for a keyword in a class

getMinFreqInClass

public int getMinFreqInClass()

The minimum frequency for a keyword in a class

withMinFreqInClass

public ClassifierConfig withMinFreqInClass(int minFreqInClass)

withMinFreqInClass

public ClassifierConfig withMinFreqInClass(java.lang.Integer minFreqInClass)

setMinScore

public void setMinScore(double minScore)

The minimum score for a keyword

getMinScore

public double getMinScore()

The minimum score for a keyword

withMinScore

public ClassifierConfig withMinScore(double minScore)

makeCopy

public ClassifierConfig makeCopy()

Creates and returns a deep copy of this ClassifierConfig.

Overrides:

makeCopy in class SupervisedLearningConfig

readFrom

public static ClassifierConfig readFrom(java.io.InputStream is)
                                 throws javax.xml.bind.JAXBException

Read this ClassifierConfig from an XML fragment.

Throws:

javax.xml.bind.JAXBException

writeTo

public void writeTo(java.io.OutputStream os)
             throws javax.xml.bind.JAXBException,
                    java.io.IOException

Write this ClassifierConfig as an XML fragment

Overrides:

writeTo in class SupervisedLearningConfig

Throws:

javax.xml.bind.JAXBException

java.io.IOException

fromString

public static ClassifierConfig fromString(java.lang.String s)
                                   throws javax.xml.bind.JAXBException,
                                          java.io.UnsupportedEncodingException

String representation of this ClassifierConfig.

Throws:

javax.xml.bind.JAXBException

java.io.UnsupportedEncodingException

toString

public java.lang.String toString()

String representation of this ClassifierConfig.

Overrides:

toString in class SupervisedLearningConfig

check

public void check(boolean deep,
                  java.lang.String errorContext)
           throws com.exalead.util.TypedException

Checks this ClassifierConfig.

Specified by:

check in interface com.exalead.util.Checkable

Overrides:

check in class SupervisedLearningConfig

Throws:

com.exalead.util.TypedException