Class RankBasedSummarizer


  • public class RankBasedSummarizer
    extends Object
    Generates summaries for single dimensions using the ranks of coherence calculations. It assumes that the data is a CSV file with the following positions:
    • 0: Probability estimation type
    • 1: Window size (or empty if no window)
    • 2: Segmentation type
    • 3: Indirect measure type (or empty)
    • 4: Gamma value of the indirect measure type (or empty)
    • 5: Direct measure type
    • 6: Aggregation type
    • 7+: Correlation values for datasets
    • last: Average correlation value
    Author:
    Michael Röder (michael.roeder@uni-paderborn.de)