For BEFAIR2 track
- Methods for detecting algorithmic discrimination
- Bias in machine learning process, including data collection, data preparation, modeling, evaluation, and deployment
- Debiasing strategies for the legal domain
- Ethical behavior from autonomous systems
- Legal design, visual law and legal knowledge base visualization
- Design justice and fairness by design
- Differences and similarities between legal and moral reasoning
- Ethics in data mining and argument mining from legal databases and texts
- Fairness analysis on computer-assisted dispute resolution
- Counterfactual reasoning for fairness
- Algorithmic fairness and learning challenges
- Demographic Parity, Equal Opportunity, Equalized Odds, Disparate Treatment,
- Counterfactual Fairness, Fairness through Awareness, and other fairness definitions
- The impossibility theorem of fairness definitions
- Measurement and mitigation of unfairness in machine learning
- Explainability, traceability, data and model lineage
- Transparency and accountability as a fundamental design requirement for AI&L architectures
For EMAI track
- Ethical issues in knowledge representation and legal reasoning
- Formal and computational models of ethical reasoning in the legal domain (e.g. using argumentation frameworks, case-based reasoning)
- Computational models of ethical decision making
- Computational models of moral reasoning
- Computational models of various ethical theories
- Moral decision-making frameworks for legal artificial agents
- Ethics, ontologies and legal knowledge representation
- Experimental implementation of moral and ethical reasoning systems into AI driven devices
- Transparency of ethical decision making
- Value-based reasoning
- Implementation of ethics in machine learning-based systems
- Deep learning and data analytics applied to ethics in the legal domain
- Integration of data- and knowledge-driven approaches to model ethical decision making
- Argumentation with values
- Ethical constraints for reinforcement learning models
- Models of evaluation of cases in the light of different values