Monday
Workshop Goals
Preferences guide human decision making from early childhood (e.g. "which ice cream flavour do you prefer?") up to complex professional and organisatorial decisions (e.g. "which investment funds to choose?"). Preferences have traditionally been studied in economics and applied to decision making problems. Recent work in AI and related fields has led to new types of preference models and new problems for applying preference structures (see, for example, the special issue on preferences of Computational Intelligence published in May 2004). The workshop promotes this broadened scope of preference handling. As such, it provides a forum for presenting advances in preference handling and for exchanging experiences between researchers facing similar questions, but coming from different fields. In particular, this workshop is intended as a multidisciplinary workshop that will bring together researchers from diverse areas, as in the recent Dagstuhl workshop on preferences held in June, 2004. It also continues the tradition of the AAAI-02 workshop on preferences which brought together two communities (NMR and CP) with fruitful results. The workshop builds on the large number of AI researchers working on preference-related issues, but aims to attract researchers from databases, multi-criteria decision making, economics, etc. This intended diversity is reflected in the program committee, and registration to this workshop does not require registration to the main conference.
Explicit preference modelling provides a declarative way to choose among alternatives, whether these are solutions of problems to solve, answers of data-base queries, decisions of a computational agent, plans of a robot, and so on. Preference-based systems allow finer-grained control over computation and new ways of interactivity, and therefore provide more satisfactory results and outcomes. Preferences are a relatively new topic to artificial intelligence and are becoming of greater interest in many areas: knowledge representation, multi-agent systems, constraint safisfaction, decision making, decision-theoretic planning, and more. Preferences are inherently a multi-disciplinary topic, of interest to economists, computer scientists, OR researchers, mathematicians and more.