Cognitively-inspired symbolic framework for knowledge representation

Abstract

© 2017 IEEE. This paper introduces a cognitively-inspired symbolic framework for knowledge representation in human-machine interaction. The framework is developed within the ongoing research on a computational model of a hierarchical associative long-term memory. The model integrates neurocognitive understanding of the human memory system with selected insights from linguistics, and primarily addresses the storage aspect of the long-term memory. The proposed memory structure is conceptualized as a set of (multisource-multisink) semantic flow networks, including knowledge units of different complexity. It also provides algorithm for semantic integration and associative learning. The model is illustrated for a dedicated interaction domain, and implemented within a prototype system.