Carlos Castillo

Carlos Castillo

Universitat Pompeu Fabra

Engineering Sciences

Carlos' goal is to address problems of social significance through computational methods and interdisciplinary research, and his current focus is on algorithmic fairness. His background is web mining and information retrieval, and he has been influential in the areas of crisis informatics and web content quality and credibility. He is a prolific, highly cited researcher who has received two test-of-time awards, five best paper awards, and two best student paper awards. His works include a book on Big Crisis Data, as well as monographs on Information and Influence Propagation, and Adversarial Web Search. He leads the Web Science and Social Computing research group at Universitat Pompeu Fabra, and coordinates the Horizon Europe FINDHR project on non-discrimination in algorithmic hiring.

Research interests

The focus of my research is algorithmic fairness. Currently, I work on automated risk assessment methods that satisfy algorithmic fairness criteria, such as separability and sufficiency, validating them through interdisciplinary research in various applied settings. I also work on algorithms for fair recommendation and fair search, seeking to give similar items similar exposure, as well as trying to steer people away from false or misleading information or harmful content.

Selected publications

- Fabbri F, Wang Y, Bonchi F, Castillo C & Mathioudakis M 2022, 'Rewiring What-to-Watch-Next Recommendations to Reduce Radicalization Pathways', Proceedings Of The ACM Web Conference 2022 (WWW'22), 2719 - 2728. WINNER, BEST PAPER AWARD.

- Karimi-Haghighi M, Castillo C & Hernandez-Leo D 2022, 'A Causal Inference Study on the Effects of First Year Workload on the Dropout Rate of Undergraduates', Artificial Intelligence In Education, 13355, 15 - 27.

- Hertweck C, Castillo C & Mathioudakis M 2022, 'Designing Affirmative Action Policies under Uncertainty', Journal of Learning Analytics, 9(2), 121-137.

- Fabbri F, Croci ML, Bonchi F & Castillo C 2022, 'Exposure Inequality in People Recommender Systems: The Long-Term Effects', Proceedings of the International AAAI Conference on Web and Social Media, vol 16, pp 194-204.

- Porcaro L, Gómez E & Castillo C 2022, 'Perceptions of Diversity in Electronic Music: the Impact of Listener, Artist, and Track Characteristics', Proceedings of the International AAAI Conference on Web and Social Media, 16(1), 194-204.

Selected research activities

I coordinated and successfully submitted a project proposal to Horizon Europe call HORIZON-CL4-2021-HUMAN-01-24: Tackling gender, race and other biases in AI. Our proposal, involving 13 partners, was named "Fairness and Intersectional Non-Discrimination in Human Recommendation" (FINDHR) and is about exploring technical, legal, and ethical aspects of algorithmic hiring.

At the end of 2022 we signed grant agreement ID: 101070212, with an EU contribution of € 3.3 Million (€ 700 K of which is for UPF). The project will take place from November 2022 through October 2025.