Enhance Knowledge Acquisition with theory Architecture

Abstract

Traditional knowledge acquisition and machine learning methods merely acquire knowledge from instances. Great amount of instances are needed for complex problems. Another problem greatly handicaps knowledge acquisition is semantic gap which is caused by lacking of knowledge about processing. Inspired by the Monte Carlo thinking and psychological facts, architecture of theories and appropriate knowledge acquisition and problem solving methods are advanced. Semantic gap and incoherence can be effectively handled with the architecture. Knowledge acquisition and problem solving can be greatly enhanced in efficiency and accuracy by implementing the architecture because of the bridging of incoherence by the theory architecture.