Package de.uni_trier.wi2.procake.similarity.nest.sequence.utils

Interface DP

All Known Subinterfaces:
DTW, SWA
All Known Implementing Classes:
DPImpl, DTWImpl, SWAImpl
public interface DP
Super class/interface for different dynamic programming approaches.

  • Method Details

    • computeSimilarity

      void computeSimilarity()
      Compute the similarity by performing dp.

    • computeSimilarity

      void computeSimilarity(NESTSequentialWorkflowObject newCaseRes)
      Compute the similarity by performing dp on the initial query tasks and new case tasks.
      Parameters:
      newCaseRes - The new case resource to compare the query to.

    • computeSimilarity

      void computeSimilarity(DataObject[] newCaseRes)
      Compute the similarity by performing dp on the initial query tasks and new case tasks.
      Parameters:
      newCaseRes - The new case resource to compare the query to.

    • getRawSimilarityScore

      double getRawSimilarityScore()
      Return the found raw similarity.
      Returns:
      The raw similarity score.

    • getNormedSimilarityScore

      double getNormedSimilarityScore()
      Return the normalized similarity.
      Returns:
      The normalized similarity score.

    • setCase

      void setCase(NESTSequentialWorkflowObject caseRes)
      Set a new case to be compared to the query.
      Parameters:
      caseRes - The new de facto instance.

    • setCase

      void setCase(DataObject[] caseRes)
      Set a new case to be compared to the query.
      Parameters:
      caseRes - The new de facto instance.

    • getAssignments

      MatrixPath getAssignments()
      Returns the found assignments of the dp algorithm.
      Returns:
      The found assignments of the dp algorithm.

    • getAlignment

      DataObject[][] getAlignment()
      Returns the found alignment of the dp algorithm as a 2d array. return[0] und return[1] enthalten die alignments.
      Returns:
      A 2d array containing the alignment.

    • getRecommendation

      DataObject getRecommendation()
      Get the recommendation resulting from the found alignment.
      Returns:
      The data object to recommend.

    • setSimilarityValuator

      void setSimilarityValuator(SimilarityValuator simValuator)
      Set a similarity valuator to be used to calculate the local values.
      Parameters:
      simValuator -

    • setHalvingDistancePercentage

      void setHalvingDistancePercentage(double halvingDistancePercentage)
      Set the desired halving distance in percentage of the entire query task length.
      Parameters:
      halvingDistancePercentage -

    • setBindToLastRow

      void setBindToLastRow(boolean b)
      Specify whether to bind alignment to last row of matrix.
      Parameters:
      b - true or false.

    • removeCase

      void removeCase()
      Remove the curernt case.

    • setLocalSimWeights

      void setLocalSimWeights(double taskSimWeight, double inputDataSimWeight, double outputDataSimWeight)
      Set the desired weight distribution for calculation of local node similarities. Only applies if de facto graphs (or lists of task nodes) are being compared.
      Parameters:
      taskSimWeight -
      inputDataSimWeight -
      outputDataSimWeight -

    • getStepVec

      Vector getStepVec(DataObject queryTask, DataObject caseTask)
      Vector representing DP-specific possibilities associated with each step. This function is implemented by DTW and SWA individually.
      Returns:
      A vector representing each step's values. (#val diagonal step#, #val horizontal step#, #val vertical step#, 0).

    • setDataSimilarityToUse

      void setDataSimilarityToUse(String name)
      Set the similarity measure to be used for data nodes during local node sim calc. Only applies if de facto graphs (or lists of task nodes) are being compared.
      Parameters:
      name - Name of the measure.

    • setLocalSimilarityToUse

      void setLocalSimilarityToUse(String name)
      Set the similarity measure to be used for task nodes during local node sim calc.
      Parameters:
      name - Name of the measure.