Summary
Exactly as humans, artificially intelligent algorithms may generalize in unacceptable ways and unintentionally discriminate certain groups. This sparks a call for deeply embedding ethical rules in data mining algorithms to guarantee fair and unbiased decision procedures. For taxation too, the fairness principle is essential and a major challenge in digitalisation. The main research question here is therefore how to implement ethical considerations in artificial intelligence taxation systems.
Research projects
Fairness in Machine Learning (project 1)
- researcher: Marco Favier
- supervisor : prof. Toon Calders
- research is funded by AXA.
- read more on this topic on the website of Antwerp Tax Academy
Fairness in Machine Learning (project 2)
- researcher: Ewoenam Topko
- supervisor : prof. Toon Calders
- research is funded by the Flemish Government
- read more on this topic on the website of Antwerp Tax Academy
Fairness in Machine Learning (project 3)
- researcher: Daphne Lenders
- supervisor : prof. Toon Calders and prof. Sylvie De Raedt
- research is funded by the University of Antwerp
- currently working (September - December 2033) on a research project on the use of explainable AI to avoid discrimination in AI models at the Scuola Normale Superiore (Pisa), where she will work with the KDD group (https://kdd.isti.cnr.it/) and under the supervision of prof. Fosca Gianotti (research stay funded by the FWO)
Publications and presentations
2023
- Tokpo Ewoenam Kwaku, Delobelle Pieter, Berendt Bettina, Calders Toon, How far can it go? On intrinsic gender bias mitigation for text classification, Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2023), May 2-6, Dubrovnik, Croatia - Association for Computational Linguistics, 2023, p. 3418-3433
- Lenders Daphne, Calders Toon, Users’ needs in interactive bias auditing tools introducing a requirement checklist and evaluating existing tools, AI Ethics, 2023
- Lenders Daphne, Calders Toon, Functional requirements for interactive bias-audit tools, HHAI 2023: Augmenting human intellect: Proceedings of the Second International Conference on Hybrid-Artificial Intelligence / Lukowicz, Paul, p. 426-428
- Lenders Daphne, Calders Toon, Real-life performance of fairness interventions : introducing a new benchmarking dataset for fair ML, Proceedings of the 38th ACM/SIGAPP Symposium on Applied Computing (SAC '23), March 27 – March 31, 2023, Tallinn, Estonia, p. 350-357
- Pinxtere Sam, Favier Marco, Calders Toon, RangeGrad : explaining neural networks by measuring uncertainty through bound propagation, Machine learning and principles and practice of knowledge discovery in databases, ECML PKDD 2022, PTI - European Conference on Machine Learning and Principles and Practice of, Knowledge Discovery in Databases (ECML PKDD), SEP 19-23, 2022, Grenoble, France, Cham :Springer international publishing ag, 2023, p. 336-352
2022
- Tokpo Ewoenam Kwaku, Calders Toon, Text style transfer for bias mitigation using masked language modeling, NAACL 2022: The 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: proceedings of the student workshop, Human Language Technologies, JUL 10-15, 2022, Seattle, WA - Stroudsburg :Assoc computational linguistics-acl, 2022, p. 163-171
- Goethals Sofie, Martens David, Calders Toon, PreCoF: counterfactual explanations for fairness, Research Square, 2022, 31 ss.
2021
Lenders Daphne en Calders Toon, Learning a fair distance function for situation testing, 21st Joint European Conference on Machine Learning and Principles and, Practice of Knowledge Discovery in Databases (ECML PKDD), SEP 13-17, 2021, Cham : Springer international publishing ag , 2021, 631-646
Pieter Delobelle, Ewoenam Kwaku Tokpo, Toon Calders, Bettina Berendt:, Measuring Fairness with Biased Rulers: A Survey on Quantifying Biases in Pretrained Language Models. CoRR abs/2112.07447 (2021)
Pinxteren Sam, Calders Toon, Efficient permutation testing for significant sequential patterns, 2021 SIAM International Conference on Data Mining (SDM), 2021, p. 9-27
- Calders Toon, Ntoutsi Eirini, Pechenizkiy Mykola, Rosenhahn Bodo, Ruggieri Salvatore, Introduction to the special section on bias and fairness in AI, SIGKDD explorations 2021, p. 1-3
2020
- Van de Vijver, Anne, Calders, Toon, Fiscale algoritmen, profilering en het recht op privéleven, Tijdschrift voor Fiscaal Recht, 2020, 611-614 (comment on the Dutch SyRI case)