Modeling the dynamics of brand opinions in information and communication interaction: analysis of the influence of values and social connections

In modeling opinion dynamics within a group, the structure of social connections is most often considered constant; however, in real groups, the exchange of opinions can affect its stability. Moreover, the choice of topics for communication within groups is not random but takes into account the value orientations of the participants. The aim of this work was to develop a platform for the quantitative description and modeling of opinion dynamics under conditions where social connections are adjusted after each act of opinion exchange, and the choice of topic for communication depends on the agents' values. In such a model, it is proposed to study the processes of information dissemination about new brands if the set of brand messages is considered as one of the topics in the information agenda, for example, in social networks and media. Methodologically, each topic in the information agenda, as well as the brand message topics, were represented as semantic vectors. Each agent was assigned a value vector, and the matching with it influenced the choice of topic for discussion. Topics semantically closer to the values were chosen more often. It is shown that in such a model, even a slight advantage in the semantic closeness of brand messages to the target group's values gives that brand a significant advantage over an alternative brand whose message is semantically farther from the values. The platform may be of interest both for fundamental research on information exchange in social networks and media and for applied tasks related to optimizing brand messages.

Keywords: agent-based modeling; opinion dynamics in a group; social network dynamics; semantic vectorization; group values; social media marketing