Arbitration: A Commutative Operator for Belief Revision.
Paolo Liberatore, Marco Schaerf · 1995
The ability of Database Systems to cope with changing situations and information coming from different sources is crucial for their applicability in real-world scenarios. Recent work in the field of belief revision and update has provided us techniques to handle change. In this paper we introduce a different form of revision aiming at capturing the process of "merging" possibly inconsistent pieces of information. We call this process arbitration. Along the lines of Gardenfors' work, we propose a set of postulates for this operator and prove a representation theorem. Keywords: knowledge representation, belief revision, non-monotonic reasoning. 1 Introduction One of the main challenges of today's software and databases systems is their ability to manage a large amount of information coming from different sources and at different moments in time. Advanced databases systems must cope with a changing world and not completely reliable sources of information by adopting a "principled" strat...