activeannoservice / annotationdefinition.generator / UpdatableAnnotationGenerator

UpdatableAnnotationGenerator

abstract class UpdatableAnnotationGenerator : AnnotationGenerator

Base class for AnnotationGenerators which are updatable, for example ML models, as compared to static models like a statistics based generator

Constructors

Name Summary

<init> Base class for AnnotationGenerators which are updatable, for example ML models, as compared to static models like a statistics based generatorUpdatableAnnotationGenerator(id: String, annotationDefinitionID: String, targetType: TargetType, name: String, description: String, input: AnnotationStepKey, finalizeCondition: FinalizeCondition, startThreshold: Int, updateThreshold: Int, dataFilter: FilterCondition, versions: MutableList<UpdatableAnnotationGeneratorVersion> = mutableListOf(), createdTimestamp: Long = System.currentTimeMillis())

Name	Summary
<init>	Base class for AnnotationGenerators which are updatable, for example ML models, as compared to static models like a statistics based generator`UpdatableAnnotationGenerator(id:` `String, annotationDefinitionID:` `String, targetType:` `TargetType, name:` `String, description:` `String, input:` `AnnotationStepKey, finalizeCondition:` `FinalizeCondition, startThreshold:` `Int, updateThreshold:` `Int, dataFilter:` `FilterCondition, versions:` `MutableList<UpdatableAnnotationGeneratorVersion> = mutableListOf(), createdTimestamp:` `Long` `= System.currentTimeMillis())`

Properties

Name	Summary
dataFilter	How to select the data for updating of this generator. Will be combined with a filter of existing finalized annotation results for the related annotation of this generator`var dataFilter:` `FilterCondition`
startThreshold	How many samples are required minimum to start training / updating / predicting.`var startThreshold:` `Int`
updateThreshold	How many new samples are required compared to the previous version to update`var updateThreshold:` `Int`
versions	`var versions:` `MutableList<UpdatableAnnotationGeneratorVersion>`

Functions

Name	Summary
optimize	Optionally, subclasses can implement optimize to optimize e.g. hyperparameters of a ML model`open suspend fun optimize(data:` `Map<Document,` `List<AnnotationResult>>):` `Unit`
update	Implement this for updating the annotation generator given the version`abstract suspend fun update(version:` `UpdatableAnnotationGeneratorVersion, data:` `Map<Document,` `List<AnnotationResult>>):` `Unit`
updateModel	Call this to update the mutable fields of an AnnotationGenerator`open fun updateModel(newAnnotationGenerator:` `AnnotationGenerator):` `Unit`

Inheritors

Name	Summary
TagSetDocumentTargetUpdatableGeneratorModel	Generic updatable AnnotationGenerator for TagSetAnnotationDefinition and DocumentTarget.`class TagSetDocumentTargetUpdatableGeneratorModel :` `UpdatableAnnotationGenerator`