Новые технологические тренды: выявление в текстах на базе использования гибридных моделей и анализа временных рядов паттернов данных
New technological trends identification is one of the most sophisticated, as well as the most important, tasks in the domain of S&T analysis. Nowadays, the leading methodologies within the domain are focused mainly on technological roadmapping, Foresight, data patterns and time series analysis, which is used to specify current and projected trends. The paper presents intelligent tools for trend identification in texts collections with hybrid approach based on the integration of classical statistical methods and the methods of information extraction. Several existing approaches are combined to be used for multilingual text collections of various genres. Ontologies driving text processing as well as documents’ characteristic vectors containing multiword terms are used. The results of statistical analysis of document collections are presented in the form of data patterns time series that are analyzed with structural methods of image analysis. OWL representation of extensional part of the trend ontological model is generated.