Formel Ontolojiler ve Betimleyici Mantıklar

Abstract

The history of ontology reveals various methodologies that examine being. Traditional ontology studies being qua being and categorizes it. Formal ontology determines the categories that are common to all entities and classifies them with formal languages using these categories as well. However, for over thirty years, formal ontologies have been studied and built outside of philosophy. The reason why ontology is separated from philosophy and becomes an interdisciplinary study is due to our need to make classifications and standardization in data and information management in accordance with reality. What lies behind this requirement is the exponential growth of data and the impossibility of analyzing it neither by humans nor by existing methods. For instance, dozens of studies are published every minute, and then scientists cannot follow all these developments with classical techniques. Therefore, the desire that machines should be involved in the processes of sharing and producing knowledge is far higher than ever. This desire, thus, has led us to develop new models for knowledge production. It has been experienced that the standardization of knowledge is essential in the construction of models; for, significant differences were found between taxonomic structures, concept definitions, and different granularity levels in the knowledge representations of various working groups, and even the integration of data was not possible. At this point, the selection, classification, and standardization of knowledge have been guaranteed by using ontological principles and methods. In this article, within information systems, we will introduce the formal ontologies, the formal theory covering the broadest knowledge of the field by selecting, classifying and standardizing it by using the principles and methods of philosophical ontologies. We will mention that we use description logics, a family of formal languages with the expressive power of natural language and high-performance reasoning algorithms, to represent this theory in machines. In essence, this article will introduce formal ontologies and description logics in the realm of knowledge representation.