Preference learning has found a lot of attention in 2012 including events at ECAI 2012 and EURO 2012. This interesting topic will be continued by a dedicated workshop DA2PL ‘from multiple criteria Decision Aid to Preference Learning’, which will be held on Nov. 15-16th at the University of Mons, Belgium.
Whereas the decision sciences have elaborated and explored preference models for explaining, predicting, and guiding decision making behaviour since more than hundred years, newer research started in the fields of artificial intelligence and other human-centered computational fields is investigating how to represent rich preference information compactly. Explicit representations of preferences permit reasoning with and about preferences and thus open up new possibilities, leading to a series of emerging research topics.
Preference learning can be considered one of them. Wikipedia defines preference learning as “a subfield in machine learning in which the goal is to learn a predictive preference model from observed preference information”. There is a dedicated workshop series on preferences learning (PL-08, PL-09, PL-10, PL-12) started in 2008 at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases. The web site on preference learning lists research groups on this topic and provides access to resources on preference learning such as software, datasets, workshops, tutorials, special issues, and books. The upcoming workshop DA2PL on ‘from multiple criteria Decision Aid to Preference Learning’ will bring together researchers from preference learning and preference modeling in order to discuss topics of common interest.
Explanations for decisions is another emerging topic, which has just started to gain the interest of researchers in preference handling. A tutorial on “Computing Explanations in Decision Making” given at the Doctoral School on Decision Theoretic AI of the ADT COST Action in Cork in 2009 and at IJCAI 2009 introduced the concept of explaining the optimality of decisions in terms of preferences based on ideas published at CSCLP 2007. Papers on explanations for decision making have, for example, been presented at ADT 2011, M-PREF 12. Those papers investigate the form and properties of explanations for particular decision methods and particular preference models.
And there are many other new topics coming up. In May 2013, there will be a workshop on ‘Conflict Resolution in Decision Making’ COREDEMA, which studies how to resolve conflicts among multiple agents having different preferences, while trying to make them all happy. The next years will certainly see further emerging research topics around the topic of preferences.
