DOI: 10.17587/prin.8.511-523

Simulation of Society Information Space Evolution Process (The Meme-Gram-Model)

A. A. Artemov, artemsince2@ya.ru, Center of Strategic Nuclear Forces Research at Academy of Military Sciences, Moscow Region, Yubileiny, 141090, Russian Federation, A. V. Galatenko, agalat@msu.ru, Mechanics and Mathematics Faculty of Lomonosov Moscow State University, Moscow, 119234, Russian Federation

Corresponding author: Artemov Artem A., Research Associate, Center of Strategic Nuclear Forces Research at Academy of Military Sciences, Moscow Region, Yubileiny, 141090, Russian Federation, E-mail: artemsince2@ya.ru

The article presents a model and an algorithm for predicting the number of copies of a meme in the future content of the societys information space (IP). Authors presents a hypothesis that the information space of contemporary society can be described by the dynamics of changing of informational messages, which are spread by the means of mass communication technologies. The prediction period of the model is one calendar week for every three values of number of meme copies in three previous weeks. The model is statistical and was made based on the analysis of changes in the content of Russian-speaking segment of information space from May to August 2014 presented by a collection of time-varying text corpuses which were generated from 300 000 most cited in social media publications of various Internet media. In the article authors evaluate the quality of the model, which show the probability of errors in the forecast of number of copies of a meme. A special feature of the algorithm is the possibility of its application for prediction of number of meme copies for any IP with similar size and diversity parameters to the Russian segment of IP. The results of the study can be summarized in the following statements:

1. based on the analysis of the changes dynamics in the number of 200 000 bigrams copies or memes (training sample), we have developed a mathematical model that solves the problem of predicting the number of meme copies in IP for the next weekly interval depending on the change in the number of copies of the meme in the 3 previous weekly intervals;
2. a theorem on the distribution of the number of meme copies in the next discrete period of time p + 1 has been produced and proved;
3. validation of the constructed statistical model of the number of meme copies distribution in the next discrete moment of time on real data showed a high accuracy of the actual values of the response within the given 95 % confidence intervals with an average error of less than 1 %.

Quality assessment of the developed model (algorithm) has been obtained as a result of conducting 100,000 experiments to forecast the number of copies of a meme in the control sample (validation set).

Keywords: Mem-Gram-Model, language model, n-gram, bigram, memetic, meme, information space, Su-John-sons distribution, distribution law, number of memes copies, statistical model, algorithm

pp. 511–523

For citation:
Artemov A. A., Galatenko A. V. Simulation of Society Information Space Evolution Process (The Meme-Gram-Model), Programmnaya Ingeneria, 2017, vol. 8, no. 11, pp. 511—523.