Search academic texts

Information × Registration Number 0225U004757, (0125U001019) , R & D reports Title Development of a transdisciplinary platform with a component architecture of cognitive services for analyzing large amounts of information popup.stage_title Створення процедур пошуку та моніторингу інформаційних ресурсів за тематичними профілями, які важливі для прийняття рішень у сфері національної безпеки та оборони. Створення процедур перетворення мережевих наративів, зміст яких впливає на процеси прийняття рішень, в інтерактивний онтологічний формат. Head Stryzhak Oleksandr Ye., д.т.н. Registration Date 18-12-2025 Organization National Center "Small Academy of Sciences of Ukraine" popup.description1 Creation of a transdisciplinary platform with a component architecture of cognitive services that implement linguistic-semantic and conceptual analysis of large amounts of information, namely narratives, documents and tabular data, identification of logistical links between them, criterion information and decision support. popup.description2 The work in the second stage solved the urgent problem regarding the need to increase the efficiency of timely, prompt receipt and semantic processing of heterogeneous information, on the basis of which the consistency and objectivity of decision-making are ensured. To solve it, a procedure for monitoring and categorizing information resources according to thematic profiles, which are important for decision-making in the field of national security and defense, and a procedure for transforming network narratives, the content of which affects decision-making processes, into an interactive ontological format have been created. Ontology-driven tools for text taxonomization for the formation of interactive knowledge bases, as well as methods of their use, have been developed. The results of the tools can be used by experts in consolidated interaction for decision-making, which increases their efficiency and reasonableness. Software and information tools for an ontology-driven linguistic corpus of scientific, technical and technological products for military purposes and tools for taxonomization of texts have been developed. An ontology-driven corpus formation method with an emphasis on ontological indexing of text fragments and a method of using ontology-driven taxonomy tools have been created. Neural network agents for the transformation of network narratives into an ontological format and means of similarity search based on linguistic statistics, distributive semantics and semantic hashing have been implemented. The integration of the obtained solutions with an expert ontology-driven decision-support platform is ensured and the possibility of their use in the field of national security and defense is demonstrated. The uniqueness of TPCACS is the presence of a recursive reducer in its architecture and its integration with ontological models and neural network agents, which enables adaptive updating of the knowledge base in response to the emergence of new data. Product Description popup.authors Stanislav O. Dovhyi Volodymyr A. Shirokov Anatoly P. Zagnitko Vitalii V. Prykhodniuk Oleksandr Y. Stryzhak Viacheslav V. Horborukov Maxim V. Nadutenko Hrihorii M. Potapov Oleh V. Franchuk Mykyta M. Yablochkov Dzhulai Anton I. Alexander I. Shkimbov popup.nrat_date 2025-12-18 Close