Social Network Analytics for Anti-Money Laundering – A Systematic Literature Review and Experimental Evaluation
_{Bruno Deprez, Toon Vanderschueren, Tim Verdonck, Bart Baesens, Wouter Verbeke [Preprint]}

The source code of the experimental evaluation of the paper Social Network Analytics for Anti-Money Laundering -- A Systematic Literature Review and Experimental Evaluation.

It provides an implementation of different network learning techniques in a uniform manner.

Data

The main experiment is done on the elliptic dataset. Additional code is present that was used to verify the correct working of the code on the Cora dataset.

This repository does not provide any data, due to size constraints. The data can be found online using the following links:

• Cora
• Elliptic
• IBM-AML

Repository structure

The structure includes folders and scripts/notebooks containing code. The files with results, i.e., csv-files, are note shown. This repository is organised as follows:

|-data
|-notebooks
    |-AnalysisResults.ipynb
|-res
|-scripts
    |-train_supervides.py
    |-test_supervised.py
    |-train_unsupervides.py
    |-test_unsupervised.py
|-src
    |-data
        |-DatasetConstruction.py
    |-methods
        |-utils
            |-decoder.py
            |-functionsNetworKit.py
            |-functionsNetworkX.py
            |-functionsTorch.py
            |-GNN.py
            |-isolation_forest.py
        |-evaluation.py
        |-experiments_supervised.py
        |-experiments_unsupervised.py
|-utils
    |-Network.py

Installing

We have provided a requirements.txt file:

pip install -r requirements.txt

Please use the above in a newly created virtual environment to avoid clashing dependencies.

Citing

Please cite our paper and/or code as follows: Use the BibTeX citation

@misc{deprez2024networkevaluation,
      title={Network Analytics for Anti-Money Laundering -- A Systematic Literature Review and Experimental Evaluation}, 
      author={Bruno Deprez and Toon Vanderschueren and Bart Baesens and Tim Verdonck and Wouter Verbeke},
      year={2024},
      eprint={2405.19383},
      archivePrefix={arXiv},
      primaryClass={cs.SI},
      url={https://arxiv.org/abs/2405.19383}, 
}