Supervision : Jean-Gabriel GANASCIA, Gauvain BOURGNE

Formalising ethical reasoning: modelling ethical processes and modelling, representing and automating causal reasoning

This thesis is in the field of computational ethics, which aims to formalise ethical reasoning. In other words, this work is part of the field that seeks to emulate our capacity as rational beings to morally evaluate an action. The formalisation of this reasoning has two objectives: to better understand it and to integrate it into computer systems to ensure that decisions made comply with chosen moral principles.
This thesis makes a contribution to the field in two ways. Firstly, it proposes a common framework for formalising faithfully the most common moral principles in Western philosophy. This first contribution can be summarised as `modelling ethical processes'. The second set of contributions pertains to the proposal for formalising causal reasoning. This formalisation not only enhances our comprehension of this reasoning but also enables its integration into computer systems, facilitating the establishment of complex causal relationships. This capability is crucial for formalising a wide range of moral principles. To ensure that our proposal can formalise all these moral principles, we have designed it to satisfy a number of conditions. Firstly, our formalisation is based on a formalism that explicitly addresses the subtleties of problems related to both causal and ethical reasoning. Secondly, our formalism's definition of causality free of any confusion with the notion of responsibility. Otherwise, it would not be common to formalise all moral principles. Finally, our proposal can handle all causal cases, including the most complex. The second group of contributions focuses on `modelling, representing and automating causal reasoning'. The main contributions of this thesis belong to this second group.

2021-2024 Publications

• 2024
  • C. Sarmiento Lozano : “Formalisation des raisonnements éthiques : modélisation des processus en éthique et modélisation, représentation et automatisation du raisonnement causal”, thesis, phd defence 04/26/2024, supervision Ganascia, Jean-Gabriel Bourgne, Gauvain (2024)
• 2023
  • Y. Munro, C. Sarmiento, I. Bloch, G. Bourgne, M.‑J. Lesot : “Dynamic Argumentation and Action Languages: Towards Explanations”, The Fourth Workshop on Explainable Logic-Based Knowledge Representation (XLoKR 2023), Rhodes, Greece (2023)
  • Y. Munro, C. Sarmiento, I. Bloch, G. Bourgne, M.‑J. Lesot : “Temporalité et causalité en argumentation abstraite”, Actes JIAF-JFPDA 2023, Strasbourg, France (2023)
• 2022
  • C. Sarmiento, G. Bourgne, K. Inoue, J.‑G. Ganascia : “Action Languages Based Actual Causality in Decision Making Contexts”, International Workshop on AI compliance mechanism (WAICOM 2022) @ 35^th International Conference on Legal Knowledge and Information Systems (JURIX 2022), Saarbrucken, Germany (2022)
  • C. Sarmiento, G. Bourgne, K. Inoue, J.‑G. Ganascia : “Action Languages Based Actual Causality in Decision Making Contexts”, The 24^th International Conference on Principles and Practice of Multi-Agent Systems, vol. 13753, Lecture Notes in Computer Science, Valence, Spain, pp. 243-259, (Springer International Publishing), (ISBN: 978-3-031-21202-4) (2022)
  • C. Sarmiento, G. Bourgne, D. Cavalli, K. Inoue, J.‑G. Ganascia : “ACTION LANGUAGES BASED ACTUAL CAUSALITY IN ETHICAL DECISION MAKING CONTEXTS”, (2022)
• 2021
  • G. Bourgne, C. Sarmiento, J.‑G. Ganascia : “ACE modular framework for computational ethics : dealing with multiple actions, concurrency and omission”, International Workshop on Computational Machine Ethics, Online event, France (2021)