A Dynamic Knowledge Generation System for Cognitive Agents

In this work we describe a knowledge based system employed for the dynamic, goal-directed and creative generation of novel knowledge in cognitive agents. This system exploits a recently introduced extension of a Description Logic of typicality (TCL) able to combine prototypical (i.e. commonsense) descriptions of concepts. In particular: given a goal expressed as a set of properties, in case an intelligent agent cannot find a concept in its knowledge base able to fulfill these properties, our system exploits the Description Logic TCL in order to find at least two concepts whose creative combination allows to satisfy the goal. The knowledge base of the agent is then extended via a mechanism of commonsense concept combination where the resulting combined concept represents the solution for the initial goal. The proposed approach has been tested in the task of object composition and compared with the responses provided by the OROC system.