A Model of Forecasting of Information Events on the Basis of the Solution of a Boundary Value Problem for Systems with Memory and Self-Organization

Abstract

Abstract—One of the problems in forecasting of news events is the development of models that allow working with a weakly structured information space of text documents. A distinctive feature of such a news space is the stochastic nature of the processes in it, the presence of memory, and the possibility of self-organization of information. It is interesting to develop a model for predicting events on the basis of a stochastic dynamics of the changing of images (or the state of the information space) of news clusters with allowance for memory and self-organization. The article discusses the schemes of transition probabilities between states in the information space, on the basis of which a nonlinear second-order differential equation is derived and a boundary value problem for predicting news events is formulated and solved. The analysis of the model described in the article shows the possibility of an increase in the probability of reaching the predicted event almost immediately after the beginning of the process of changing the structure of news clusters and the presence of abrupt jumps and oscillations in the probability of reaching an event.