Modelling of modern social behavior in presence of cyber social networks

Abstract

This thesis contains a comprehensive study of modelling modern social behavior in presence of cyber social networks. Modelling approach used in this study is cellular automata (CA), which is a discrete model study in complexity science. Study is mainly focused on finding any deviation on typical CA model of social dynamic when impact of cyber social networks included. First half of study focuses on modelling up to date study of social dynamics and discussion of sociological theories to verify this modelling approach. Model contains most important aspect of social process that is opinion changing process which is also called opinion dynamics. Demonstrate and simulate CA models open source toolkit (Cellular Automata Visualization Program - CAVP) of cellular automata with visualization functionalities has enhanced and extended. Almost all aspects discussed in this study has been graphically simulated using CAVP toolkit. Second half of the study is focused on including cyber social network impact of influence over typical cellular automata model of social dynamics. These social networks appear as virtual groups over typical population. There are three types of social network users according to their opinion changing process. They are Stubbornness users, Compromise Users and Biased Conformity Users. Stubbornness users do not change their opinion according to the social network opinion. This study has comprehensively analyzed how other two types of users change their opinion according to the impact of social networks. Simulation results shows that compromising users will get assigned stable majority opinion and they if the assigned opinion is opposed to their cellular opinion these users toggle between network opinion and cellular opinion when they belong to strong clusters. This shows that social network users who experience majority of the member in their social network group opinion tends to hold a particular opinion they start conflicting with cellular opinion. On the other hand biased conformity users though they do not belongs to strong clusters they accept the most weighted opinion of network and toggle back to cellular opinion. Therefore, these users are the most affected by social networks and they are very unstable with their opinion changing process. To get more accurate simulations results, the size of the social network group has been considered and simulation has been carried out for small networks size to large networks and all the values are randomly generated and assigned.