Measuring Similarity of Observations Made by Artificial Cognitive Agents

A model for artificial cognitive agent has been considered in which the agent can carry out autonomous observations of an external world and stores them in the so called base profiles. For this model four general approaches to defining similarity of stored observations have been proposed and a model for their integration into one computational mechanism has been presented. For each strategy an example similarity function has been given, fulfilling some commonsense expectations defined for strategies of comparing the content of individual observations.