Selective information retrieval from hierarchical associative knowledge learning memory

For the purpose of building an efficient information system in a dynamic environment, it is necessary to develop a well structured intelligent system which has the mechanism of automatic knowledge acquisition, inference and extraction. It must also have a selective mechanism which can retrieve the information by the selecting factor as much as a user wants, because a user doesn´t need all the associated data in the memory. We propose a hierarchical associative system with selective information retrieval mechanism considering the above functions. We applied this system to estimating the purchasing degree from a customer´s tastes, the pattern of commodities and evaluation of a company