Semantic Communication With Conceptual Spaces

Despite the fact that Shannon and Weaver's Mathematical Theory of Communication was published over 70 years ago, all communication systems continue to operate at the first of three levels defined in this theory: the technical level. In this letter, we argue that a transition to the semantic level embodies a natural, important step in the evolution of communication technologies. Furthermore, we propose a novel approach to engineering semantic communication using conceptual spaces and functional compression. We introduce a model of semantic communication utilizing this approach, and simulate communication of image semantics demonstrating a 99.79% reduction in rate.

Automated summary

Despite being published over 70 years ago, Shannon and Weaver's Mathematical Theory of Communication still applies to all communication systems at the technical level. In this letter, we argue for the significance of transitioning to the semantic level as an important step in the evolution of communication technologies. We propose a novel approach using conceptual spaces and functional compression to engineer semantic communication, demonstrating a 99.79% rate reduction in simulating image semantics.