Технологии и принципы сбора и обработки неструктурированных распределенных данных с учетом современных особенностей предоставления медиа-контента

Abstract

Purpose The COVID-19 pandemic is creating serious challenges for modern society that leads to develop new information models and methods of digital monitoring not only of the spread of the virus, but also of the socio-economic environment. Materials and methods: As sources for clarifying the parameters of such models, it is advisable to choose not a limited set of predefi ned Internet sources, but unstructured media data on an unlimited set of resources, which leads to the need to build a system for complex monitoring of social phenomena. Such system can supplement and correct mathematical and information models for the spread of viruses, aimed at minimizing the damage caused by any pandemic. Results: It is proposed to create a software system that includes a Data Retrieving subsystem (for collecting and preprocessing media data) combined with a headless browser. This allows to build a system for monitoring of social phenomena, complementing mathematical and information models of the spread of viruses, aimed at minimizing the damage they cause. The feature of developed system is the using of a natural language processing framework based on the associative-ontological approach, and software implementation of the adaptive-behavioral SEIR model, as well as a subsystem for interpreting the collected data, generating metadata for identifying and correcting the model. Conclusions: The proposed system allows to make more balanced management decisions based on the analysis of the current situation in the infosphere. An additional advantage of the system is the ability to identify poorly predictable reactions of society to certain events expressed in media content.