The concept of text processing in an ontological approach to spatio-temporal social network analysis

The task of analyzing the spatio-temporal properties of social network objects and predicting behavioral characteristics based on them requires the development of new methods and approaches to managing knowledge extracted from heterogeneous information sources. This study analyses the most popular systems for processing textual arrays and forming ontologies to integrate heterogeneous information resources and shows that existing technologies do not allow structural reflection of multi-thematic knowledge due to the multiplicity and disorderliness of their semantic properties, as well as the uncertainty of their descriptions. To integrate heterogeneous data and ensure their semantic interoperability, a multilevel ontology is proposed, based on the concept of text processing, which focuses on the possibility of synchronizing text content from different social networks with information stored in ontologies.