Towards A Unified Framework for Learning and Processing Perceptual, Relational, and Meta Knowledge
Marc Pickett · 2013
We present a framework for representing, learning, and processing meta-knowledge. Our framework leverages a system initially developed for perceptual learning and processing to process both relational structures in general and feature hierarchies in particular. We describe preliminary results demonstrating how our system can be used to learn meta-ontologies, or feature hierarchies of feature hierarchies.