Checking the Operability of the Social Media Collection and Predictive Analysis System

As part of the research, a system of collecting and predictive analysis of social media was implemented. The results of predictive analytics can be used in the following cases:• When creating recommendation services for interested users in a particular product/service or for analyzing data with a small number of characteristics for clustering.• When predicting accidents/failures – for example, for early repair of equipment to reduce the risks of downtime in production.• When making the best decisions in a given situation.During the development of the system, data collection modules from VK, YouTube and OK were implemented. In addition, an analysis module has been implemented based on the repeated application of the MeanShift unsupervised learning method with cutting off the number of analyzed resources. Based on the results of the analysis, the user is offered a list of those resources on which information on the entered keywords may appear in the future. Thanks to the developed algorithm of work, resources can be divided into "most suitable", "suitable", "possibly suitable" depending on the clustering step.This work is devoted to checking the operability of the system. Tests are given to evaluate the result of predictive analysis depending on various queries for data collection.