Prospects for Predicting Coordinated Actions by Large Groups of Users Based on the Hierarchy of Activity in Social Networks
摘要
The ability to predict the actions of social network users is relevant for almost all commercial enterprises. It depends on the accuracy of the forecast, whether users click an advertisement link, buy a product or read an article. Due to the network effect, these processes sometimes develop as an avalanche and spill out beyond the Internet. There is always a symmetrical response to these processes; positive initiatives are supported (volunteering, disaster relief) and negative problems are solved (riots in the streets). Forecasting such phenomena will allow us to take timely measures and avoid the negative consequences of the acute phase of mass conflicts. The article examines the indicators of users’ activity in social networks and introduces their hierarchy based on the time spent on each type of activity. The contributing factors and the response of regulatory authorities are considered using the example of behavior of visitors to the Reddit social network and the impact of their activities on the share price of GameStop (GME shares).