Communication in multi-agent systems (MASs) is usually governed by agent communication languages (ACLs) and communication
protocols carrying a clear cut semantics. With an increasing degree of
openness, however, the need arises for more flexible models of communication that can handle the uncertainty associated with the fact
that adherence to a supposedly agreed specification of possible conversations cannot be ensured on the side of other agents.
In this paper, we argue for adaptiveness in agent communication. We present a particular approach that combines conversation patterns as a generic way of describing the available means of communication in a MAS with a decisiontheoretic framework and various
different machine learning techniques for applying these patterns in and adapting them from actual conversations.