Andrew Garland & Richard Alterman
It is often possible to envision the ways in which knowledge will be reused. By preparing the knowledge appropriately at storage time, we can simplify the later task of adapting the stored knowledge. Preparation can simplify, remove inefficiencies, and segment the trace data into useful ideas. In this paper, we apply this principle to a computational model that deals with the problem of a distributed collective memory for multi-agent systems and we provide technical detail about how the experience of the agents is prepared before storage.