19^th International CODATA Conference
Category: Infoscience

Toward a Formal Multi-aspect Description of Knowledge Resources' Structure

Prof., D.Sc. Juliusz L. Kulikowski (jlkulik@ibib.waw.pl)
Institute of Biocybernetics and Biomedical Engineering PAS, Poland

The problem of formal knowledge structure description belongs to the most important ones from the philosophical, psychological and practical points of view. In philosophy the structure of human knowledge and its relation to the real world is a basic Gnostic problem. In psychology the problem is related to this of the mechanisms of human thinking. In practice the problem is strongly connected with the methodology of knowledge bases design, maintaining and exploration. This paper will be focused on the last aspects of the problem. Therefore, we will be talking about knowledge resources structure rather than about this of knowledge itself. Knowledge resource will be considered as a enumerable set of knowledge items, each knowledge item being a proposition having one of the following standard form: 1/ factographical statements: "A is (does, has the form, property, value, etc.) B"; 2/ inference statements: "If P then Q" , 3/ causal statements: "Q, because P", 4/ commands: "Do P in order to Q", P and Q being, as before, some factographical statements. The above-described four types of statements make possible to express a large variety of facts, phenomena, processes, rules, programs, etc. However, on the higher level of a knowledge resource structure various types of relationships between the knowledge items occur. For example, some factographical statements may be collected into subsets as concerning the same subject A. Some pairs of factographical statements can be combined as satisfying the inference or causal statements, etc. So, on the second structural level a set of relationships among the statements can be established. This structure can be extended on the third and higher levels of knowledge resource's description. The relations can describe various aspects of knowledge structure: partition into subject areas and subareas, as well as the methodological, organizational, legal, historical, etc. aspects. It will be shown that an extended algebra of relations is a flexible tool that can be used in formal multi-level knowledge resources' structure modeling.