Package opennlp.tools.ml.libsvm.doccat

Class SvmDoccatConfiguration

java.lang.Object
opennlp.tools.ml.libsvm.doccat.SvmDoccatConfiguration
All Implemented Interfaces:
Serializable
public class SvmDoccatConfiguration extends Object implements Serializable
Configuration for SVM-based document categorization, combining the underlying SVM classifier settings with text-specific parameters for term weighting, feature selection, and feature scaling.
See Also:

  • Method Details

    • getSvmConfiguration

      public de.hhn.mi.configuration.SvmConfiguration getSvmConfiguration()
      Returns:
      The underlying SVM configuration for the classifier/trainer.

    • getTermWeightingStrategy

      public TermWeightingStrategy getTermWeightingStrategy()
      Returns:
      The term weighting strategy used to compute feature values.

    • getFeatureSelectionStrategy

      public FeatureSelectionStrategy getFeatureSelectionStrategy()
      Returns:
      The feature selection strategy used to reduce the feature space.

    • getMaxFeatures

      public int getMaxFeatures()
      Returns:
      The maximum number of features to select. A value <= 0 means all features are retained.

    • isScaleFeatures

      public boolean isScaleFeatures()
      Returns:
      true if feature values should be scaled to the configured range.

    • getScaleLower

      public double getScaleLower()
      Returns:
      The lower bound of the scaling range (default: 0.0).

    • getScaleUpper

      public double getScaleUpper()
      Returns:
      The upper bound of the scaling range (default: 1.0).