Intelligent trend-mining as one of the contemporary fields of linguistic research

Intelligent trend-mining (automated trend detection) from unstructured textual information flows is essential for forecasting and strategic planning. To date, research in the field, due to the complexity of deep automated text analysis, is represented by a set of highly specialized methods and tools. Currently there is no methodological basis or software for them to be ported to new domains and languages. This article provides an overview of the most representative works in the field of intellectual trend-mining and attempts to systematize the main approaches and methods to solve particular problems arising in the development of intellectual trend-mining technologies, the ultimate goal of which is processing changes in text collections, meaningful interpretation of the detected changes and generation of “readable” reports. Summarized are the main stages of intellectual text analysis when extracting trends. Interrelation between trend-mining and content analysis is underlined. Special emphasis is made on the need to develop language-independent trend-mining techniques, specialized ontologies being recognized as a valuable resource to achieve this goal. The most relevant techniques for trend-mining report generation are considered.