On the semantics of merging

Abstract

Intelligent agents are often faced with the problem of trying to merge possibly conflicting pieces of information obtained from different sources into a consistent view of the world. We propose a framework for the modelling of such merging operations with roots in the work of Spohn (1988, 1991). Unlike most approaches we focus on the merging of epistemic states, not knowledge bases. We construct a number of plausible merging operations and measure them against various properties that merging operations ought to satisfy. Finally, we discuss the connection between merging and the use of infobases Meyer (1999) and Meyer et al. (2000).