AUTONOMOUS LEARNING-TYPE KNOWLEDGE CONSTRUCTION MACHINE

摘要

An autonomous learning-type knowledge construction machine learns human thought patterns from input information and the like and generates a thought pattern similar to a series of thought patterns produced by a human with respect to a certain situation. A pattern corresponding to the human thought patterns has a connection relationship set to a pattern corresponding to a thought pattern subsequently conceived, and a significant connection relationship is constructed between patterns that are recorded. Based on knowledge that has already been recorded, the value of input information has its logical validity evaluated, and the information is updated to a more logically valid logic. With regard to information that is difficult to understand, search for additional information is conducted to expand knowledge autonomously. While in the initial steps, knowledge, relationships between knowledge, solutions to problems and the like are taught by humans and the like, in the next steps, the value of information is autonomously determined and the machine constructs knowledge by itself to construct knowledge similar to that of humans and even knowledge beyond that of humans, whereby accurate situational judgement and behavior can be conducted.