Final stages of designing a system for collecting and predictive analysis of social media data

Abstract

This article discusses the design of a system for collecting and predictive analysis of social media data. With the development of the Internet, as well as social media, it has become easier to access and distribute information because the network users themselves are both creators and recipients of varying information. The main type of social media is social networks. Facebook, VK, Instagram, YouTube, Twitter, Odnoklassniki, WhatsApp and Telegram messengers are among the most well-known ones. The most important functions of social media are to influence the perception, attitude and final behavior of consumers. Predictive analytics is based on automatic search for connections, anomalies and patterns between various factors. To form a predictive model, a large set of statistical modeling methods, data mining, machine learning, neural networks and other mechanisms are used. Together with various methods of collecting information from Internet resources, such as parsing and social network APIs, predictive analytics can offer the most interesting sources of information for the user. In order to combine the methods of predictive analysis and data collection methods, it is necessary to take a detailed approach to the system design process. In this paper, special attention is paid to the detailed description of the second of the main parts of the system (the analysis subsystem). In addition, the full architecture and algorithm of operation are highlighted. The results obtained will be used in further development, and it is planned to use them in full. Working on this topic will facilitate the process of subsequent testing and research of the system.