This article represents approaches of the artificial intelligence methods, used in public administration. An overview of various technologies of artificial intelligence applications in the field of public administration and related fields is given. All of these research directions are particularly relevant to the task of digital technologies (including artificial intelligence) growth to create an efficient and competitive digital economics in Russia. Among the modern intellectual technologies that allow solving the widest range of tasks, an important role plays technologies related to natural language text processing - nonstructured NL-texts are essential segment of data, used for analysis and decision-making tasks (in terms of data volume, of course, video data have a leading position, however, they are usually suitable for solving tactical but not strategic management tasks). The article provides an overview of the existing methods of natural language processing and their practical application to the tasks of public administration. An integrated approach to the natural language processing tools using for solving practical problems in the field of public administration is considered on the example of the ISIDA-T system for extracting information from natural language texts, developed at the Artificial Intelligence Research Center PSI RAS. The system under consideration is distinguished by a modular approach to the pre-processing of unstructured text and the possibility of manual adjustment for a specific extraction task. This technological solution gives the necessary flexibility and ease of use. The system consists of configurable text preprocessing, linguistic analysis, target information retrieval and output of the results in user-friendly form modules. One of the important components of the system is an integrated knowledge resource that allows quickly and efficiently adjust the system to the specifics of the relevant subject area. An approach to the application of the ISIDA-T system to the task of fact information extraction (as the most important for analyzing the situation and subsequent management decisions) is proposed using the example of information extraction about resignations and appointments from news feeds. © 2019 Copyright is held by the owner/author(s). Publication rights licensed to ACM.