Train Classifier

Verb: trainClassifier

Available from: <Enterprise>

Trains a text classifier to update its database information with a new version, according to the classification algorithm selected, and separates categories by grouping text files by folders.

To train the text classifier when updating your database, it is first necessary to create a model in "Tools> Machine Learning Model Builder".

Syntax

trainClassifier --folder(String) --algorithm(Nullable<MachineLearningAlgorithm>) --culture(Culture) --name(String) [--production(Boolean)]

Inputs

Script Designer Required AcceptedTypes Description
--folder Directory Required Text Path where the content used to train the classifier is located.
--algorithm Algorithm Required MachineLearningAlgorithm Algorithm used to categorize the content of Directory. Options:
  • Bag-of-Words
  • Functional Bag-of-Words
  • Functional N-Gram
  • Functional Text Classifier
  • Knowledge Base V2
  • N-Gram
  • Text Classifier
    		•
    --culture Language Required Culture Language in which the contents of Directory and the text classifier database are written. Options:
  • en-US
  • pt-BR
  • pt-PT
    		•
    --name Name Required Text Name of the text classifier database to be trained.
    --production Mark as Production Optional Boolean When enabled, it allows the version of the trained classifier to be used in a production environment.

    See Also

  • Train Knowledge Base