Extended visit to the University of Pisa and Rome in 2024. Looking at two main aspects:
- Theoretical/Ethical – Theory and modelling relating to privacy, bias and explainability. “The aim will be to document a semi-formal exploration of emerging human issues, in particular, using new algorithmic developments as case studies or inspirations to highlight or challenge more conceptual understanding.”
- Technical – Developing interaction and infrastructure for working with large data in ways that can enhance user understanding, whilst protecting personal data sovereignty. “The visit will allow further exploration of these issues and early steps to create better means for user interaction with big data that takes into account both the human capabilities of the data users and human rights of the data subjects.”
Working documents
- [WD1] Being Secure – some myths of privacy and personal security
- [WD2] Interesting Latent Spaces
- [WD3] Models of Trust and Reliance
- [WD4] Two way communication and local human explanations
- [WD5] A third way – XAI betwixt symbolic and sub-symbolic
- [WD6] QbB as informal case study / worked example
Talks and papers arising from the visit
- Alan Dix (2024). Patient Interaction – for well-being, productivity and sustainability. FUSION 2024, Kuala Lumpur, Malaysia, 28 Sept. 2024. detailed notes and slides
- Alan Dix (2024). Whose choice? Making decisions with and about Artificial Intelligence. HCI-E 2024, Istanbul, Turkey, 22 Nov. 2024. abstract and notes
- A. Dix, T. Turchi, B. Wilson, A. Monreale and M. Roach. (2025). Talking Back – human input and explanations to interactive AI systems. Workshop on Adaptive eXplainable AI (AXAI), IUI 2025, Cagliari, Italy, 24th March 2025. abstract and paper
- A. Dix. (2024). AI Changes Everything. Talk at the final event of Data Fusion Dynamics: A Collaborative UK-Saudi Initiative in Cybersecurity and Artificial Intelligence funded by the British Council UK-Saudi Challenge Fund 2024, Cardiff Metropolitan University, 29th April 2025 detailed notes and slides
- Alan Dix (2024). Whose choice? Making decisions with and about Artificial Intelligence. Invited talk at Designing for People: AI and the Benefits of Human-Centred Digital Products, Digital & AI Revolution Week, Keele University, 14th May 2025 abstract and slides
- A. Dix, T. Turchi, B. Wilson, A. Monreale and M. Roach. (2025). Maintaining Coherence in Explainable AI: Strategies for Consistency Across Time and Interaction. SYNERGY – Designing and Building Hybrid Human–AI Systems || Workshop on Adaptive eXplainable AI (AXAI), HHAI 2025, Pisa, Italy, 9th June 2025. full paper
Other references
- [Dx90] A. J. Dix (1990). Information processing, context and privacy. Human-Computer Interaction – INTERACT'90, Ed. D. G. D. Diaper G. Cockton & B. Shakel. North-Holland. pp. 15-20. https://alandix.com/academic/papers/int90/
- [Dx19] A. Dix. (2019). Being secure: some myths of privacy and personal security. Talk at Cardiff University, 6th February 2019. https://alandix.com/academic/talks/being-secure-2019/
- [Dxp24] Alan Dix (2024). Patient Interaction – for well-being, productivity and sustainability. FUSION 2024, Kuala Lumpur, Malaysia, 28 Sept. 2024. https://www.alandix.com/academic/talks/ FUSION2024/
- [dJ83] Wiebren de Jonge, "Compromising statistical databases responding to queries about means", ACM Trans. On Database Systems 8(1), pp. 60-80, ACM (March 1983). https://doi.org/10.1145/319830.319834
- [NM22] Naretto, F., Monreale, A., and Giannotti, F. (2022). Evaluating the privacy exposure of interpretable global explainers. In 2022 IEEE 4th International Conference on Cognitive Machine Intelligence (CogMI) (pp. 13-19). IEEE. https://doi.org/10.1109/CogMI56440.2022.00012
|
Cluster-centroid latent space
[zoom image]
Modelling a single decision where the human uses AI assistance
[zoom image]
Clamping concept nodes in a deep neural network
[zoom image]
primary/sensed and secondary/supported tasks
mediated by data
[zoom image]