Class/Object

com.johnsnowlabs.nlp.annotators.ner.crf

NerCrfApproach

Related Docs: object NerCrfApproach | package crf

Permalink

class NerCrfApproach extends AnnotatorApproach[NerCrfModel] with NerApproach[NerCrfApproach]

Algorithm for training a Named Entity Recognition Model

For instantiated/pretrained models, see NerCrfModel.

This Named Entity recognition annotator allows for a generic model to be trained by utilizing a CRF machine learning algorithm. The training data should be a labeled Spark Dataset, e.g. CoNLL 2003 IOB with Annotation type columns. The data should have columns of type DOCUMENT, TOKEN, POS, WORD_EMBEDDINGS and an additional label column of annotator type NAMED_ENTITY. Excluding the label, this can be done with for example

Optionally the user can provide an entity dictionary file with setExternalFeatures for better accuracy.

For extended examples of usage, see the Spark NLP Workshop and the NerCrfApproachTestSpec.

Example

import com.johnsnowlabs.nlp.base.DocumentAssembler
import com.johnsnowlabs.nlp.annotators.Tokenizer
import com.johnsnowlabs.nlp.annotators.sbd.pragmatic.SentenceDetector
import com.johnsnowlabs.nlp.embeddings.WordEmbeddingsModel
import com.johnsnowlabs.nlp.annotators.pos.perceptron.PerceptronModel
import com.johnsnowlabs.nlp.training.CoNLL
import com.johnsnowlabs.nlp.annotator.NerCrfApproach
import org.apache.spark.ml.Pipeline

// This CoNLL dataset already includes a sentence, token, POS tags and label
// column with their respective annotator types. If a custom dataset is used,
// these need to be defined with for example:

val documentAssembler = new DocumentAssembler()
  .setInputCol("text")
  .setOutputCol("document")

val sentence = new SentenceDetector()
  .setInputCols("document")
  .setOutputCol("sentence")

val tokenizer = new Tokenizer()
  .setInputCols("sentence")
  .setOutputCol("token")

val posTagger = PerceptronModel.pretrained()
  .setInputCols("sentence", "token")
  .setOutputCol("pos")

// Then the training can start
val embeddings = WordEmbeddingsModel.pretrained()
  .setInputCols("sentence", "token")
  .setOutputCol("embeddings")
  .setCaseSensitive(false)

val nerTagger = new NerCrfApproach()
  .setInputCols("sentence", "token", "pos", "embeddings")
  .setLabelColumn("label")
  .setMinEpochs(1)
  .setMaxEpochs(3)
  .setOutputCol("ner")

val pipeline = new Pipeline().setStages(Array(
  embeddings,
  nerTagger
))

// We use the sentences, tokens, POS tags and labels from the CoNLL dataset.
val conll = CoNLL()
val trainingData = conll.readDataset(spark, "src/test/resources/conll2003/eng.train")

val pipelineModel = pipeline.fit(trainingData)
See also

NerConverter to further process the results

NerDLApproach for a deep learning based approach

Linear Supertypes
NerApproach[NerCrfApproach], AnnotatorApproach[NerCrfModel], CanBeLazy, DefaultParamsWritable, MLWritable, HasOutputAnnotatorType, HasOutputAnnotationCol, HasInputAnnotationCols, Estimator[NerCrfModel], PipelineStage, Logging, Params, Serializable, Serializable, Identifiable, AnyRef, Any
Ordering
  1. Grouped
  2. Alphabetic
  3. By Inheritance
Inherited
  1. NerCrfApproach
  2. NerApproach
  3. AnnotatorApproach
  4. CanBeLazy
  5. DefaultParamsWritable
  6. MLWritable
  7. HasOutputAnnotatorType
  8. HasOutputAnnotationCol
  9. HasInputAnnotationCols
  10. Estimator
  11. PipelineStage
  12. Logging
  13. Params
  14. Serializable
  15. Serializable
  16. Identifiable
  17. AnyRef
  18. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Instance Constructors

  1. new NerCrfApproach()

    Permalink
  2. new NerCrfApproach(uid: String)

    Permalink

    uid

    required uid for storing annotator to disk

Type Members

  1. type AnnotatorType = String

    Permalink
    Definition Classes
    HasOutputAnnotatorType

Value Members

  1. final def !=(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  2. final def ##(): Int

    Permalink
    Definition Classes
    AnyRef → Any
  3. final def $[T](param: Param[T]): T

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  4. final def ==(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  5. def _fit(dataset: Dataset[_], recursiveStages: Option[PipelineModel]): NerCrfModel

    Permalink
    Attributes
    protected
    Definition Classes
    AnnotatorApproach
  6. final def asInstanceOf[T0]: T0

    Permalink
    Definition Classes
    Any
  7. def beforeTraining(spark: SparkSession): Unit

    Permalink
    Definition Classes
    AnnotatorApproach
  8. val c0: IntParam

    Permalink

    c0 params defining decay speed for gradient (Default: 2250000)

  9. final def checkSchema(schema: StructType, inputAnnotatorType: String): Boolean

    Permalink
    Attributes
    protected
    Definition Classes
    HasInputAnnotationCols
  10. final def clear(param: Param[_]): NerCrfApproach.this.type

    Permalink
    Definition Classes
    Params
  11. def clone(): AnyRef

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  12. final def copy(extra: ParamMap): Estimator[NerCrfModel]

    Permalink
    Definition Classes
    AnnotatorApproach → Estimator → PipelineStage → Params
  13. def copyValues[T <: Params](to: T, extra: ParamMap): T

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  14. final def defaultCopy[T <: Params](extra: ParamMap): T

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  15. val description: String

    Permalink

    CRF based Named Entity Recognition Tagger

    CRF based Named Entity Recognition Tagger

    Definition Classes
    NerCrfApproachAnnotatorApproach
  16. val entities: StringArrayParam

    Permalink

    Entities to recognize

    Entities to recognize

    Definition Classes
    NerApproach
  17. final def eq(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  18. def equals(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  19. def explainParam(param: Param[_]): String

    Permalink
    Definition Classes
    Params
  20. def explainParams(): String

    Permalink
    Definition Classes
    Params
  21. val externalFeatures: ExternalResourceParam

    Permalink

    Additional dictionary to use for features

  22. final def extractParamMap(): ParamMap

    Permalink
    Definition Classes
    Params
  23. final def extractParamMap(extra: ParamMap): ParamMap

    Permalink
    Definition Classes
    Params
  24. def finalize(): Unit

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( classOf[java.lang.Throwable] )
  25. final def fit(dataset: Dataset[_]): NerCrfModel

    Permalink
    Definition Classes
    AnnotatorApproach → Estimator
  26. def fit(dataset: Dataset[_], paramMaps: Array[ParamMap]): Seq[NerCrfModel]

    Permalink
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" )
  27. def fit(dataset: Dataset[_], paramMap: ParamMap): NerCrfModel

    Permalink
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" )
  28. def fit(dataset: Dataset[_], firstParamPair: ParamPair[_], otherParamPairs: ParamPair[_]*): NerCrfModel

    Permalink
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" ) @varargs()
  29. final def get[T](param: Param[T]): Option[T]

    Permalink
    Definition Classes
    Params
  30. def getC0: Int

    Permalink

    c0 params defining decay speed for gradient

  31. final def getClass(): Class[_]

    Permalink
    Definition Classes
    AnyRef → Any
  32. final def getDefault[T](param: Param[T]): Option[T]

    Permalink
    Definition Classes
    Params
  33. def getIncludeConfidence: Boolean

    Permalink

    Whether or not to calculate prediction confidence by token, includes in metadata

  34. def getInputCols: Array[String]

    Permalink

    returns

    input annotations columns currently used

    Definition Classes
    HasInputAnnotationCols
  35. def getL2: Double

    Permalink

    L2 regularization coefficient

  36. def getLazyAnnotator: Boolean

    Permalink
    Definition Classes
    CanBeLazy
  37. def getLossEps: Double

    Permalink

    If Epoch relative improvement less than eps then training is stopped

  38. def getMaxEpochs: Int

    Permalink

    Maximum number of epochs to train

    Maximum number of epochs to train

    Definition Classes
    NerApproach
  39. def getMinEpochs: Int

    Permalink

    Minimum number of epochs to train

    Minimum number of epochs to train

    Definition Classes
    NerApproach
  40. def getMinW: Double

    Permalink

    Features with less weights then this param value will be filtered

  41. final def getOrDefault[T](param: Param[T]): T

    Permalink
    Definition Classes
    Params
  42. final def getOutputCol: String

    Permalink

    Gets annotation column name going to generate

    Gets annotation column name going to generate

    Definition Classes
    HasOutputAnnotationCol
  43. def getParam(paramName: String): Param[Any]

    Permalink
    Definition Classes
    Params
  44. def getRandomSeed: Int

    Permalink

    Random seed

    Random seed

    Definition Classes
    NerApproach
  45. def getVerbose: Int

    Permalink

    Level of verbosity during training

    Level of verbosity during training

    Definition Classes
    NerApproach
  46. final def hasDefault[T](param: Param[T]): Boolean

    Permalink
    Definition Classes
    Params
  47. def hasParam(paramName: String): Boolean

    Permalink
    Definition Classes
    Params
  48. def hashCode(): Int

    Permalink
    Definition Classes
    AnyRef → Any
  49. val includeConfidence: BooleanParam

    Permalink

    Whether or not to calculate prediction confidence by token, included in metadata (Default: false)

  50. def initializeLogIfNecessary(isInterpreter: Boolean, silent: Boolean): Boolean

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  51. def initializeLogIfNecessary(isInterpreter: Boolean): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  52. val inputAnnotatorTypes: Array[String]

    Permalink

    Input annotator types : DOCUMENT, TOKEN, POS, WORD_EMBEDDINGS

    Input annotator types : DOCUMENT, TOKEN, POS, WORD_EMBEDDINGS

    Definition Classes
    NerCrfApproachHasInputAnnotationCols
  53. final val inputCols: StringArrayParam

    Permalink

    columns that contain annotations necessary to run this annotator AnnotatorType is used both as input and output columns if not specified

    columns that contain annotations necessary to run this annotator AnnotatorType is used both as input and output columns if not specified

    Attributes
    protected
    Definition Classes
    HasInputAnnotationCols
  54. final def isDefined(param: Param[_]): Boolean

    Permalink
    Definition Classes
    Params
  55. final def isInstanceOf[T0]: Boolean

    Permalink
    Definition Classes
    Any
  56. final def isSet(param: Param[_]): Boolean

    Permalink
    Definition Classes
    Params
  57. def isTraceEnabled(): Boolean

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  58. val l2: DoubleParam

    Permalink

    L2 regularization coefficient (Default: 1f)

  59. val labelColumn: Param[String]

    Permalink

    Column with label per each token

    Column with label per each token

    Definition Classes
    NerApproach
  60. val lazyAnnotator: BooleanParam

    Permalink
    Definition Classes
    CanBeLazy
  61. def log: Logger

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  62. def logDebug(msg: ⇒ String, throwable: Throwable): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  63. def logDebug(msg: ⇒ String): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  64. def logError(msg: ⇒ String, throwable: Throwable): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  65. def logError(msg: ⇒ String): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  66. def logInfo(msg: ⇒ String, throwable: Throwable): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  67. def logInfo(msg: ⇒ String): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  68. def logName: String

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  69. def logTrace(msg: ⇒ String, throwable: Throwable): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  70. def logTrace(msg: ⇒ String): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  71. def logWarning(msg: ⇒ String, throwable: Throwable): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  72. def logWarning(msg: ⇒ String): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Logging
  73. val lossEps: DoubleParam

    Permalink

    If Epoch relative improvement is less than lossEps then training is stopped (Default: 1e-3f)

  74. val maxEpochs: IntParam

    Permalink

    Maximum number of epochs to train

    Maximum number of epochs to train

    Definition Classes
    NerApproach
  75. val minEpochs: IntParam

    Permalink

    Minimum number of epochs to train

    Minimum number of epochs to train

    Definition Classes
    NerApproach
  76. val minW: DoubleParam

    Permalink

    Features with less weights then this param value will be filtered

  77. def msgHelper(schema: StructType): String

    Permalink
    Attributes
    protected
    Definition Classes
    HasInputAnnotationCols
  78. final def ne(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  79. final def notify(): Unit

    Permalink
    Definition Classes
    AnyRef
  80. final def notifyAll(): Unit

    Permalink
    Definition Classes
    AnyRef
  81. def onTrained(model: NerCrfModel, spark: SparkSession): Unit

    Permalink
    Definition Classes
    AnnotatorApproach
  82. val optionalInputAnnotatorTypes: Array[String]

    Permalink
    Definition Classes
    HasInputAnnotationCols
  83. val outputAnnotatorType: AnnotatorType

    Permalink

    Output annotator types : NAMED_ENTITY

    Output annotator types : NAMED_ENTITY

    Definition Classes
    NerCrfApproachHasOutputAnnotatorType
  84. final val outputCol: Param[String]

    Permalink
    Attributes
    protected
    Definition Classes
    HasOutputAnnotationCol
  85. lazy val params: Array[Param[_]]

    Permalink
    Definition Classes
    Params
  86. val randomSeed: IntParam

    Permalink

    Random seed

    Random seed

    Definition Classes
    NerApproach
  87. def save(path: String): Unit

    Permalink
    Definition Classes
    MLWritable
    Annotations
    @Since( "1.6.0" ) @throws( ... )
  88. final def set(paramPair: ParamPair[_]): NerCrfApproach.this.type

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  89. final def set(param: String, value: Any): NerCrfApproach.this.type

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  90. final def set[T](param: Param[T], value: T): NerCrfApproach.this.type

    Permalink
    Definition Classes
    Params
  91. def setC0(c0: Int): NerCrfApproach.this.type

    Permalink

    c0 params defining decay speed for gradient

  92. final def setDefault(paramPairs: ParamPair[_]*): NerCrfApproach.this.type

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  93. final def setDefault[T](param: Param[T], value: T): NerCrfApproach.this.type

    Permalink
    Attributes
    protected
    Definition Classes
    Params
  94. def setEntities(tags: Array[String]): NerCrfApproach

    Permalink

    Entities to recognize

    Entities to recognize

    Definition Classes
    NerApproach
  95. def setExternalFeatures(path: String, delimiter: String, readAs: Format = ReadAs.TEXT, options: Map[String, String] = Map("format" -> "text")): NerCrfApproach.this.type

    Permalink

    Additional dictionary to use for features

  96. def setExternalFeatures(value: ExternalResource): NerCrfApproach.this.type

    Permalink

    Additional dictionary to use for features

  97. def setIncludeConfidence(c: Boolean): NerCrfApproach.this.type

    Permalink

    Whether or not to calculate prediction confidence by token, includes in metadata

  98. final def setInputCols(value: String*): NerCrfApproach.this.type

    Permalink
    Definition Classes
    HasInputAnnotationCols
  99. final def setInputCols(value: Array[String]): NerCrfApproach.this.type

    Permalink

    Overrides required annotators column if different than default

    Overrides required annotators column if different than default

    Definition Classes
    HasInputAnnotationCols
  100. def setL2(l2: Double): NerCrfApproach.this.type

    Permalink

    L2 regularization coefficient

  101. def setLabelColumn(column: String): NerCrfApproach

    Permalink

    Column with label per each token

    Column with label per each token

    Definition Classes
    NerApproach
  102. def setLazyAnnotator(value: Boolean): NerCrfApproach.this.type

    Permalink
    Definition Classes
    CanBeLazy
  103. def setLossEps(eps: Double): NerCrfApproach.this.type

    Permalink

    If Epoch relative improvement less than eps then training is stopped

  104. def setMaxEpochs(epochs: Int): NerCrfApproach

    Permalink

    Maximum number of epochs to train

    Maximum number of epochs to train

    Definition Classes
    NerApproach
  105. def setMinEpochs(epochs: Int): NerCrfApproach

    Permalink

    Minimum number of epochs to train

    Minimum number of epochs to train

    Definition Classes
    NerApproach
  106. def setMinW(w: Double): NerCrfApproach.this.type

    Permalink

    Features with less weights then this param value will be filtered

  107. final def setOutputCol(value: String): NerCrfApproach.this.type

    Permalink

    Overrides annotation column name when transforming

    Overrides annotation column name when transforming

    Definition Classes
    HasOutputAnnotationCol
  108. def setRandomSeed(seed: Int): NerCrfApproach

    Permalink

    Random seed

    Random seed

    Definition Classes
    NerApproach
  109. def setVerbose(verbose: Level): NerCrfApproach

    Permalink

    Level of verbosity during training

    Level of verbosity during training

    Definition Classes
    NerApproach
  110. def setVerbose(verbose: Int): NerCrfApproach

    Permalink

    Level of verbosity during training

    Level of verbosity during training

    Definition Classes
    NerApproach
  111. final def synchronized[T0](arg0: ⇒ T0): T0

    Permalink
    Definition Classes
    AnyRef
  112. def toString(): String

    Permalink
    Definition Classes
    Identifiable → AnyRef → Any
  113. def train(dataset: Dataset[_], recursivePipeline: Option[PipelineModel]): NerCrfModel

    Permalink
    Definition Classes
    NerCrfApproachAnnotatorApproach
  114. final def transformSchema(schema: StructType): StructType

    Permalink

    requirement for pipeline transformation validation.

    requirement for pipeline transformation validation. It is called on fit()

    Definition Classes
    AnnotatorApproach → PipelineStage
  115. def transformSchema(schema: StructType, logging: Boolean): StructType

    Permalink
    Attributes
    protected
    Definition Classes
    PipelineStage
    Annotations
    @DeveloperApi()
  116. val uid: String

    Permalink

    required uid for storing annotator to disk

    required uid for storing annotator to disk

    Definition Classes
    NerCrfApproach → Identifiable
  117. def validate(schema: StructType): Boolean

    Permalink

    takes a Dataset and checks to see if all the required annotation types are present.

    takes a Dataset and checks to see if all the required annotation types are present.

    schema

    to be validated

    returns

    True if all the required types are present, else false

    Attributes
    protected
    Definition Classes
    AnnotatorApproach
  118. val verbose: IntParam

    Permalink

    Level of verbosity during training (Default: Verbose.Silent.id)

    Level of verbosity during training (Default: Verbose.Silent.id)

    Definition Classes
    NerApproach
  119. final def wait(): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  120. final def wait(arg0: Long, arg1: Int): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  121. final def wait(arg0: Long): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  122. def write: MLWriter

    Permalink
    Definition Classes
    DefaultParamsWritable → MLWritable

Inherited from NerApproach[NerCrfApproach]

Inherited from AnnotatorApproach[NerCrfModel]

Inherited from CanBeLazy

Inherited from DefaultParamsWritable

Inherited from MLWritable

Inherited from HasOutputAnnotatorType

Inherited from HasOutputAnnotationCol

Inherited from HasInputAnnotationCols

Inherited from Estimator[NerCrfModel]

Inherited from PipelineStage

Inherited from Logging

Inherited from Params

Inherited from Serializable

Inherited from Serializable

Inherited from Identifiable

Inherited from AnyRef

Inherited from Any

Parameters

A list of (hyper-)parameter keys this annotator can take. Users can set and get the parameter values through setters and getters, respectively.

Annotator types

Required input and expected output annotator types

Members

Parameter setters

Parameter getters