SEMANTIC CORE OF DIGITAL PLATFORM

Abstract

This article gives a generalized description of the integrated complex of declarative and procedural means, which together provide a consistent information (semantic) representation of complex objects at all stages of their life cycle. Knowledge (facts) is represented by a system of ontologies of different levels, and the structure - by taxonomies and classifications. Ontology is defined as a system of three interrelated systems (functional, conceptual and sign), and the system of taxonomies represents classes of objects and processes’ characteristic of the main "coordinates" of activity. Information retrieval, as a component of the semantic core, is considered as a complex self-consistent process of constructing new knowledge, where knowledge is information (texts of found documents) associated with the context of the task and the user's views. Such context is purposefully or indirectly determined by the user by means of predefined semantic structures (taxonomies, ontologies) or by means of dynamically formed components (dictionaries, samples, etc.). This is the essence of the semantic cognitive search, where the system not only implements the selection of documents by traditional retrieval methods, but also forms an image of information needs, which, in turn, will allow the system to synthesize complex, aspect-oriented answers. Automated technologies of linguistic support based on distributivestatistical analysis of both object knowledge and implicit knowledge flows (extracted by the system in the interaction process) and components of conceptual-terminological systems are proposed. Scientific and educational potential can be widely used to create a basic component of information and linguistic support in the early stages. The tools presented in the article are tested in the framework of the developed software xIrbis-ML, designed for semantic search of data in knowledge bases for complex engineering objects.