Benchmark of ARX’s Heuristic Algorithms

Benchmark of ARX's heuristic algorithms and further approaches for anonymizing high-dimensional data.

Workflow

Benchmark:
The Java code contained in this repository can be used to run the benchmarks. The used datasets and the corresponding generalization hierarchies are located in the folders data and hierarchies. The benchmark results are provided as CSV-files and can be found in the results folder.

Visualization:
The Python code can be used to analyze and visualize the CSV-files contained in results. The figures generated by the script can be found in the figures folder.

Contents

The content of each sub-directory is briefly described below. Additionally, each subdirectory contains its own readme file for further information.

data: Datasets that were used as input for the experiments. README
hierarchies: Generalization hierarchies used for the experiments to specify how attribute values can be altered README
java: Java code required to run the anonymization experiments. README
results: Results obtained from running the experiments. README
python: Python code used to analyze and visualize the results. README
figures: Figures obtained from visualizing the results of the experiments. README