Bibliographische Detailangaben

Beteiligte Personen:	Taki, Hirokazu (VerfasserIn), Terasaki, Satoshi (VerfasserIn)
Format:	Buch
Sprache:	Englisch
Veröffentlicht:	Tokyo, Japan 1990
Schriftenreihe:	Shin-Sedai-Konpyūta-Gijutsu-Kaihatsu-Kikō <Tōkyō>: ICOT technical report 546
Schlagwörter:	Knowledge acquisition (Expert systems) Machine learning
Abstract:	Abstract: "This paper describes a knowledge acquisition approach which is an integration of an interactive knowledge acquisition and an inductive knowledge generation. In this approach, problem solving knowledge is extracted as a set of operations. There are three knowledge acquisition phases in the knowledge acquisition process. The first is the expert model construction phase, in which problem solving knowledge is extracted as a set of operations and the type of each operation is defined. Operation types are similar to operations of relational algebra. The second phase is the model instantiation phase, in which detailed knowledge is extracted according to the operation types Sometimes, it is not easy to extract detailed information without proposals of knowledge. Therefore, this approach supports an inductive mechanism which makes operation presumptions in order to stimulate experts to associate detailed knowledge. The third phase is the knowledge refinement phase, in which the knowledge base is evaluated at the operation level to check whether the knowledge is sufficient. This paper also introduces an interactive knowledge acquisition support system, EPSILON/One, which has an interactive knowledge acquisition function and operation presumption algorithms based on induction.
Umfang:	28 S.