Building cognitive knowledge bases sharable by humans and cognitive robots

Abstract

One of the fundamental means for interaction and coordination between humans and cognitive robots is knowledge sharing in general and formal concept comprehension in particular. A cognitive knowledge base (CKB) is introduced as a formal structure of collective knowledge embodied by a weighted hierarchical concept network. This paper formally describes a CKB based on concept algebra and semantic algebra. An Algorithm of CKB Generation (ACKBG) is developed for autonomous machine learning from complex human knowledge and semantic expressions. A set of experiments demonstrates applications of ACKBG in brain-machine interfacing via a sharable CKB in cognitive computing, semantic computing, and machine learning.