At present, ontologies are recognized as the most effective means of formalizing and systematizing knowledge and data in scientific subject domains (SSDs). However, the development of an ontology is a rather complicated and time-consuming process. All indications are that when developing SSDs ontologies, it is especially effective to use ontology design patterns (ODPs). This is due to the fact that the SSD ontology, as a rule, contains a large number of typical frag-ments, which are well described by the ODPs. In addition, due to the fact that the use of ODPs greatly facilitates the development of an SSD ontology, it is possible to involve experts in a modeled SSD not possessing the skills of onto-logical modeling. To obtain an ontology that adequately describes the SSD, it is necessary to process a huge number of publications relevant to the modeled SSD. It is possible to facilitate and accelerate the process of populating the ontolo-gy with information from such sources by using the lexical and syntactic patterns of ontological design. The paper pre-sents an approach to the automated development of SSDs ontologies based on a system of heterogeneous ODPs. This system includes both ODPs intended for ontology developers and lexical and syntactic patterns built on the basis of the above-mentioned types of the ODPs and the current version of the SSD ontology.
APPROACHES TO KNOWLEDGE EXTRACTION IN SCIENTIFIC SUBJECT DOMAINS
