Interface DependencyRetriever

  • All Superinterfaces:
    Retriever<DataObject,​DependencyQuery>
    All Known Implementing Classes:
    DependencyRetrieverImpl

    public interface DependencyRetriever
    extends Retriever<DataObject,​DependencyQuery>
    Interface for the dependency specific retriever, that implements the approach of dependency-guided retrieval. This follows a brute force approach like the LinearRetrieverImpl, but deals with a dependency specific query, consisting of several data objects, and outputs a dependency specific retrieval result list with several elements. For more information, see paper.
    Author:
    Rahol Kumar, Alexander Schultheis
    • Field Detail

      • DEFAULT_RESULT_NUMBER

        static final int DEFAULT_RESULT_NUMBER
        The default value for the result number is 10. In the paper, this value is referred to as k.
        See Also:
        Constant Field Values
      • DEFAULT_ALPHA

        static final double DEFAULT_ALPHA
        The default value for the weighting parameter alpha, that enables adjusting both parts of the similarity computation to individual needs of the use case. The default value is 0.5. For this value, both, the dependency similarity and the computed similarity are weighted equally.
        See Also:
        Constant Field Values
    • Method Detail

      • setResultsNumber

        void setResultsNumber​(int resultsNumber)
        Method, to set the parameter k for the result numbers, that defines, how many dependency full matches must be found. If not at least k such case combinations were detected, a search of combined cases using dependency similarities is performed.
        Parameters:
        resultsNumber - The parameter k defines, how many dependency full matches must be found.
      • getResultsNumber

        int getResultsNumber()
        Method, that returns the parameter k for the result numbers, that defines, how many dependency full matches must be found. If not at least k such case combinations were detected, a search of combined cases using dependency similarities is performed.
        Returns:
        The parameter k defines, how many dependency full matches must be found.
      • setAlpha

        void setAlpha​(double alpha)
        Method, to set the parameter alpha, which is a weighting factor, that enables adjusting both parts of the similarity computation to individual needs of the use case. The default weight is DEFAULT_ALPHA.
        Parameters:
        alpha - The weighting parameter alpha, that should be used.
      • getAlpha

        double getAlpha()
        Method, that returns the parameter alpha, which is a weighting factor, that enables adjusting both parts of the similarity computation to individual needs of the use case.
        Returns:
        The weighting parameter alpha, that is used.