Development of a method for extracting spatial data from texts for visualization and information decision-making support for territorial management

Abstract

Abstract Information support for decision making for planning activities in the Arctic and management of the Arctic territories of the Russian Federation using spatial information visualization tools is an important and urgent task. To effectively address this challenge, it is necessary to develop specialized information tools that focus on the analysis, processing, and visualization of large amounts of information to support decision-making. Such tools include methods and technologies for extracting spatial data from texts in natural language for their further visualization. This paper describes the first part of a large complex of research works aimed at creating methods, technologies, algorithms and software for visualization of spatial data extracted from texts in natural language. At this stage of the work the technology of geodata extraction from the Arctic texts written in Russian is being developed. The paper also substantiates the prospects of application of the created methods and tools to support decision making in the field of territory planning and regional management. The review of named entities recognition problems on the basis of the analysis of scientific works executed in various areas of processing and the analysis of texts is resulted. The main approaches to text analysis and the most significant results achieved so far in this area are considered. The technology developed by the authors to extract spatial data from texts to support decision-making in planning and management of territories is part of a larger system to support socio-economic development of the region. This system is primarily focused on the Arctic territories of the Russian Federation and includes modules for analysis and visualization of geodata identified in the text analysis. The paper describes the research problem, the methods and tools used for text analysis, as well as the main results.