Packages

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 Examples 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

NerDLApproach for a deep learning based approach

NerConverter to further process the results

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()
  2. new NerCrfApproach(uid: String)

    uid

    required uid for storing annotator to disk

Type Members

  1. type AnnotatorType = String
    Definition Classes
    HasOutputAnnotatorType

Value Members

  1. final def !=(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  2. final def ##(): Int
    Definition Classes
    AnyRef → Any
  3. final def $[T](param: Param[T]): T
    Attributes
    protected
    Definition Classes
    Params
  4. final def ==(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  5. def _fit(dataset: Dataset[_], recursiveStages: Option[PipelineModel]): NerCrfModel
    Attributes
    protected
    Definition Classes
    AnnotatorApproach
  6. final def asInstanceOf[T0]: T0
    Definition Classes
    Any
  7. def beforeTraining(spark: SparkSession): Unit
    Definition Classes
    AnnotatorApproach
  8. val c0: IntParam

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

  9. final def checkSchema(schema: StructType, inputAnnotatorType: String): Boolean
    Attributes
    protected
    Definition Classes
    HasInputAnnotationCols
  10. final def clear(param: Param[_]): NerCrfApproach.this.type
    Definition Classes
    Params
  11. def clone(): AnyRef
    Attributes
    protected[lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( ... ) @native()
  12. final def copy(extra: ParamMap): Estimator[NerCrfModel]
    Definition Classes
    AnnotatorApproach → Estimator → PipelineStage → Params
  13. def copyValues[T <: Params](to: T, extra: ParamMap): T
    Attributes
    protected
    Definition Classes
    Params
  14. final def defaultCopy[T <: Params](extra: ParamMap): T
    Attributes
    protected
    Definition Classes
    Params
  15. val description: String

    CRF based Named Entity Recognition Tagger

    CRF based Named Entity Recognition Tagger

    Definition Classes
    NerCrfApproach → AnnotatorApproach
  16. val entities: StringArrayParam

    Entities to recognize

    Entities to recognize

    Definition Classes
    NerApproach
  17. final def eq(arg0: AnyRef): Boolean
    Definition Classes
    AnyRef
  18. def equals(arg0: Any): Boolean
    Definition Classes
    AnyRef → Any
  19. def explainParam(param: Param[_]): String
    Definition Classes
    Params
  20. def explainParams(): String
    Definition Classes
    Params
  21. val externalFeatures: ExternalResourceParam

    Additional dictionary to use for features

  22. final def extractParamMap(): ParamMap
    Definition Classes
    Params
  23. final def extractParamMap(extra: ParamMap): ParamMap
    Definition Classes
    Params
  24. def finalize(): Unit
    Attributes
    protected[lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( classOf[java.lang.Throwable] )
  25. final def fit(dataset: Dataset[_]): NerCrfModel
    Definition Classes
    AnnotatorApproach → Estimator
  26. def fit(dataset: Dataset[_], paramMaps: Seq[ParamMap]): Seq[NerCrfModel]
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" )
  27. def fit(dataset: Dataset[_], paramMap: ParamMap): NerCrfModel
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" )
  28. def fit(dataset: Dataset[_], firstParamPair: ParamPair[_], otherParamPairs: ParamPair[_]*): NerCrfModel
    Definition Classes
    Estimator
    Annotations
    @Since( "2.0.0" ) @varargs()
  29. final def get[T](param: Param[T]): Option[T]
    Definition Classes
    Params
  30. def getC0: Int

    c0 params defining decay speed for gradient

  31. final def getClass(): Class[_]
    Definition Classes
    AnyRef → Any
    Annotations
    @native()
  32. final def getDefault[T](param: Param[T]): Option[T]
    Definition Classes
    Params
  33. def getIncludeConfidence: Boolean

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

  34. def getInputCols: Array[String]

    returns

    input annotations columns currently used

    Definition Classes
    HasInputAnnotationCols
  35. def getL2: Double

    L2 regularization coefficient

  36. def getLazyAnnotator: Boolean
    Definition Classes
    CanBeLazy
  37. def getLossEps: Double

    If Epoch relative improvement less than eps then training is stopped

  38. def getMaxEpochs: Int

    Maximum number of epochs to train

    Maximum number of epochs to train

    Definition Classes
    NerApproach
  39. def getMinEpochs: Int

    Minimum number of epochs to train

    Minimum number of epochs to train

    Definition Classes
    NerApproach
  40. def getMinW: Double

    Features with less weights then this param value will be filtered

  41. final def getOrDefault[T](param: Param[T]): T
    Definition Classes
    Params
  42. final def getOutputCol: String

    Gets annotation column name going to generate

    Gets annotation column name going to generate

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

    Random seed

    Random seed

    Definition Classes
    NerApproach
  45. final def hasDefault[T](param: Param[T]): Boolean
    Definition Classes
    Params
  46. def hasParam(paramName: String): Boolean
    Definition Classes
    Params
  47. def hashCode(): Int
    Definition Classes
    AnyRef → Any
    Annotations
    @native()
  48. val includeConfidence: BooleanParam

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

  49. def initializeLogIfNecessary(isInterpreter: Boolean, silent: Boolean): Boolean
    Attributes
    protected
    Definition Classes
    Logging
  50. def initializeLogIfNecessary(isInterpreter: Boolean): Unit
    Attributes
    protected
    Definition Classes
    Logging
  51. val inputAnnotatorTypes: Array[String]

    Input annotator types : DOCUMENT, TOKEN, POS, WORD_EMBEDDINGS

    Input annotator types : DOCUMENT, TOKEN, POS, WORD_EMBEDDINGS

    Definition Classes
    NerCrfApproach → HasInputAnnotationCols
  52. final val inputCols: StringArrayParam

    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
  53. final def isDefined(param: Param[_]): Boolean
    Definition Classes
    Params
  54. final def isInstanceOf[T0]: Boolean
    Definition Classes
    Any
  55. final def isSet(param: Param[_]): Boolean
    Definition Classes
    Params
  56. def isTraceEnabled(): Boolean
    Attributes
    protected
    Definition Classes
    Logging
  57. val l2: DoubleParam

    L2 regularization coefficient (Default: 1f)

  58. val labelColumn: Param[String]

    Column with label per each token

    Column with label per each token

    Definition Classes
    NerApproach
  59. val lazyAnnotator: BooleanParam
    Definition Classes
    CanBeLazy
  60. def log: Logger
    Attributes
    protected
    Definition Classes
    Logging
  61. def logDebug(msg: ⇒ String, throwable: Throwable): Unit
    Attributes
    protected
    Definition Classes
    Logging
  62. def logDebug(msg: ⇒ String): Unit
    Attributes
    protected
    Definition Classes
    Logging
  63. def logError(msg: ⇒ String, throwable: Throwable): Unit
    Attributes
    protected
    Definition Classes
    Logging
  64. def logError(msg: ⇒ String): Unit
    Attributes
    protected
    Definition Classes
    Logging
  65. def logInfo(msg: ⇒ String, throwable: Throwable): Unit
    Attributes
    protected
    Definition Classes
    Logging
  66. def logInfo(msg: ⇒ String): Unit
    Attributes
    protected
    Definition Classes
    Logging
  67. def logName: String
    Attributes
    protected
    Definition Classes
    Logging
  68. def logTrace(msg: ⇒ String, throwable: Throwable): Unit
    Attributes
    protected
    Definition Classes
    Logging
  69. def logTrace(msg: ⇒ String): Unit
    Attributes
    protected
    Definition Classes
    Logging
  70. def logWarning(msg: ⇒ String, throwable: Throwable): Unit
    Attributes
    protected
    Definition Classes
    Logging
  71. def logWarning(msg: ⇒ String): Unit
    Attributes
    protected
    Definition Classes
    Logging
  72. val lossEps: DoubleParam

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

  73. val maxEpochs: IntParam

    Maximum number of epochs to train

    Maximum number of epochs to train

    Definition Classes
    NerApproach
  74. val minEpochs: IntParam

    Minimum number of epochs to train

    Minimum number of epochs to train

    Definition Classes
    NerApproach
  75. val minW: DoubleParam

    Features with less weights then this param value will be filtered

  76. def msgHelper(schema: StructType): String
    Attributes
    protected
    Definition Classes
    HasInputAnnotationCols
  77. final def ne(arg0: AnyRef): Boolean
    Definition Classes
    AnyRef
  78. final def notify(): Unit
    Definition Classes
    AnyRef
    Annotations
    @native()
  79. final def notifyAll(): Unit
    Definition Classes
    AnyRef
    Annotations
    @native()
  80. def onTrained(model: NerCrfModel, spark: SparkSession): Unit
    Definition Classes
    AnnotatorApproach
  81. val optionalInputAnnotatorTypes: Array[String]
    Definition Classes
    HasInputAnnotationCols
  82. val outputAnnotatorType: AnnotatorType

    Output annotator types : NAMED_ENTITY

    Output annotator types : NAMED_ENTITY

    Definition Classes
    NerCrfApproach → HasOutputAnnotatorType
  83. final val outputCol: Param[String]
    Attributes
    protected
    Definition Classes
    HasOutputAnnotationCol
  84. lazy val params: Array[Param[_]]
    Definition Classes
    Params
  85. val randomSeed: IntParam

    Random seed

    Random seed

    Definition Classes
    NerApproach
  86. def save(path: String): Unit
    Definition Classes
    MLWritable
    Annotations
    @Since( "1.6.0" ) @throws( ... )
  87. final def set(paramPair: ParamPair[_]): NerCrfApproach.this.type
    Attributes
    protected
    Definition Classes
    Params
  88. final def set(param: String, value: Any): NerCrfApproach.this.type
    Attributes
    protected
    Definition Classes
    Params
  89. final def set[T](param: Param[T], value: T): NerCrfApproach.this.type
    Definition Classes
    Params
  90. def setC0(c0: Int): NerCrfApproach.this.type

    c0 params defining decay speed for gradient

  91. final def setDefault(paramPairs: ParamPair[_]*): NerCrfApproach.this.type
    Attributes
    protected
    Definition Classes
    Params
  92. final def setDefault[T](param: Param[T], value: T): NerCrfApproach.this.type
    Attributes
    protected[org.apache.spark.ml]
    Definition Classes
    Params
  93. def setEntities(tags: Array[String]): NerCrfApproach

    Entities to recognize

    Entities to recognize

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

    Additional dictionary to use for features

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

    Additional dictionary to use for features

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

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

  97. final def setInputCols(value: String*): NerCrfApproach.this.type
    Definition Classes
    HasInputAnnotationCols
  98. def setInputCols(value: Array[String]): NerCrfApproach.this.type

    Overrides required annotators column if different than default

    Overrides required annotators column if different than default

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

    L2 regularization coefficient

  100. def setLabelColumn(column: String): NerCrfApproach

    Column with label per each token

    Column with label per each token

    Definition Classes
    NerApproach
  101. def setLazyAnnotator(value: Boolean): NerCrfApproach.this.type
    Definition Classes
    CanBeLazy
  102. def setLossEps(eps: Double): NerCrfApproach.this.type

    If Epoch relative improvement less than eps then training is stopped

  103. def setMaxEpochs(epochs: Int): NerCrfApproach

    Maximum number of epochs to train

    Maximum number of epochs to train

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

    Minimum number of epochs to train

    Minimum number of epochs to train

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

    Features with less weights then this param value will be filtered

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

    Overrides annotation column name when transforming

    Overrides annotation column name when transforming

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

    Random seed

    Random seed

    Definition Classes
    NerApproach
  108. def setVerbose(verbose: Level): NerCrfApproach.this.type

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

  109. def setVerbose(verbose: Int): NerCrfApproach.this.type

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

  110. final def synchronized[T0](arg0: ⇒ T0): T0
    Definition Classes
    AnyRef
  111. def toString(): String
    Definition Classes
    Identifiable → AnyRef → Any
  112. def train(dataset: Dataset[_], recursivePipeline: Option[PipelineModel]): NerCrfModel
    Definition Classes
    NerCrfApproach → AnnotatorApproach
  113. final def transformSchema(schema: StructType): StructType

    requirement for pipeline transformation validation.

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

    Definition Classes
    AnnotatorApproach → PipelineStage
  114. def transformSchema(schema: StructType, logging: Boolean): StructType
    Attributes
    protected
    Definition Classes
    PipelineStage
    Annotations
    @DeveloperApi()
  115. val uid: String
    Definition Classes
    NerCrfApproach → Identifiable
  116. def validate(schema: StructType): Boolean

    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
  117. val verbose: IntParam

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

  118. final def wait(): Unit
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  119. final def wait(arg0: Long, arg1: Int): Unit
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  120. final def wait(arg0: Long): Unit
    Definition Classes
    AnyRef
    Annotations
    @throws( ... ) @native()
  121. def write: MLWriter
    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