A.S. Zagarskikh - Engineer, National Research University of Information Technologies, Mechanics and Optics
A.S. Karsakov - Engineer, National Research University of Information Technologies, Mechanics and Optics
S.V. Ivanov - Ph.D. (Eng.), Senior Research Scientist, National Research University of Information Technologies, Mechanics and Optics
A.V. Boukhanovsky - Dr. Sc. (Eng.), Professor, National Research University of Information Technologies, Mechanics and Optics

Management of large urban (urbanized) areas (LUA) is a complex problem which requires a comprehensive review of the various urban subsystems. To provide decision support, including the rapid operative reaction and to plan the territories - development, a variety of Geographic Information Systems (GIS) together with data from regular statistical sources (e.g., regional Information Analysis Centers), as well as open data of social media, are traditionally used. Aggregation, processing and analyses of such data are performed by the "big data" model, and the calculation of case scenarios requires the use of supercomputing technology, which considerably limits the use of traditional GIS solutions and development tools. All mentioned above makes the creation of instrumental shells (frameworks) to simplify the development and implementation of such systems in different situations an actual problem. This article presents a framework, that implements a GIS technology and allows the developers to build completed decision support systems (DSS) with the use of flexible high-level interface, providing the ability to interpret the socio-economic characteristics of the LUA and their representation in the form of interactive maps and cognitive images for decision making. Apart from traditional sources of statistical information, a new generation of DSS to manage LUA is focused on open source social media (Vkontakte, Twitter, Foursquare, Instagram, Livejournal), that provides an API for easy data retrieval. A distributed crawler of social networks, which supports a traditional big data model of distributed computing MapReduce based on Hadoop environment, collects data according to their geolocation. To interpret the results of data analysis and modeling, the GIS technology based on large-format interactive table that provides the collaborative decision support for a group of experts is applied. This tool sets the requirements and layout of GIS interface, which defines how to work with DSS in general, and its architecture. As a consequence, it can put together all basic functions of such environment within the framework, which allows the experts to provide the fundamental functionality of visualization, control, access and interpretation of geo-data and as well as of computer modeling using the remote computing resources. The implementation of GIS technology in this framework is carried out by the subsystem of layers. All layers have a unified interface that simplifies the work with layers for the system as a whole. The subsystem interface to a distributed computing environment based on cloud platform CLAVIRE provides a seamless integration of separate software modules, and each of these modules works on its own computing resources. This allows the computational package that operate on remote supercomputers to connect to the DSS. The presented framework provides an instrumental base for rapid development of advanced decision support system (DSS) to manage LUA, which is focused on the shared use of cloud computing technologies, resource-intensive modeling and GIS technology.