This article describes the main properties of an iterative method of simple knowledge structure creation. The method is based on an inductive learning scheme. The knowledge structure is built automatically and takes the form of a simplified ontology. Knowledge transformation plays a key role in the process of creating the knowledge structure. In order to regular describe many kinds of these transformations the article provides the relevant theoretical background. The task of finding the proper ontology (knowledge structure) is extremely complex. This paper highlights the necessity to investigate efficient search methods; additionally, the work draws attention to the advantages that arise from building the knowledge structure at the minimal possible size. The paper points to possible areas of the method application, especially in connection with problems of the automatic understanding of images and websites.
