Topics

The workshop investigates the new reasoning and problem solving capabilities of explicit preference models for all relevant subfields of AI, including multi-agent systems, planning and robotics, vision and perception, natural language processing, knowledge representation and reasoning, constraint satisfaction and search, cognitive modeling and human interaction, and for AI-related fields such as social choice and consensus methods. As most problems studied in Artificial Intelligence involve some form of choices, they are potential applications for AI-based preference handling methods. Preferences can be applied to

• conflicting goals and alternative actions in planning and robotics
• alternative interpretations in perception and natural language understanding
• conflicting beliefs in belief revision
• alternative hypotheses in diagnostic reasoning
• choices in combinatorial problems such as configuration, design, and auctions
• agent choices in multi-agent systems
• group choices
• choices in reasoning and search strategies

Of particular interest are new emerging questions concerning the potential of preferences for AI problems:

• What are the benefits of new capabilities such as reasoning about preferences and revision of preferences?
• Can preferences help to reason about a huge space of alternatives?
• Can preferences coordinate choices in multiple simultaneous tasks such as perception, reasoning, and action?
• How to find consensus among votes in topics such as multi-agent systems, data mining, and ontology formation (such as phylogenetic trees)?

The workshop will also continue the innovation of new preference handling methods based on ideas from Artificial Intelligence. Topics such as knowledge representation for preferences, complexity and expressiveness of preference representations, reasoning about preferences, preference learning, and preference change are of ongoing interest for the workshop.