DEVS modelling and simulation of human social interaction and influence

Abstract

The social influence is at the centre of consideration in social science. In industrial engineering, although the enterprise has reached the age of the electronic communication, the human direct communication is not sufficiently considered even if it remains critical communication vector to transmit information. The idea is to predict some human attributes behavior that will help enterprise to make efficient decision. The research in the domain gives significant results but the impact of information on individuals within a social network is, mostly, statically modelled where the dynamic aspect is not frequently tackled. The individual's reaction to a change within an organization or ecosystem (implementation of a new system, new security instructions...etc.) is not always rationale. The opinion of individuals is influenced by information gathered about the attributes of the technology from other members of their social network. In addition, the works about modelling and simulation of the population's reactions to an event do not use explicit specification languages to support their models. A behavioural specification model is one critical missing link. Adding a clear behavioural model can help for specification verification and reuse. From literature, the DEVS formalism (Discrete EVent system Specifications) appears being general enough to represent such dynamical systems (Zeigler et al., 2000). It provides operational semantics applicable to this domain. The contributions of this work are dynamic models of individuals using low-level language to simulate the propagation of information among a group of individuals and its influence on their behaviour. In more details, we define a set of models of individuals characterized by a set of state variables and the mesh between the individuals within a social network. Then, we introduce the information diffusion based on epidemic spreading algorithms and we transpose them into the case of the message propagation in a social network. Finally, a basic scenario is used to give a beginning of validation to our models using a platform based on DEVS formalism.

Cell_DEVS: An Approach to Model the Influence of Social Interactions in Human Behavior

“…In this paper, we will present a mechanism for studying the evolution of human opinion due to social interaction and influence using M&S. We will discuss different models of human social interaction based on the model presented by (Bouanan, Zacharewicz, and Vallespir 2016). Our goal is to provide a serious game, so for example, marketing managers can understand how human social interactions can influence others opinion about a new product launch.…”

“…people with a specific social status or an age range) may influence the opinion of other groups or individuals towards the acquisition of a product or the adoption of a service. Based on the model presented by (Bouanan, Zacharewicz, and Vallespir 2016), we also use Cell-DEVS formalism, Agent-Based Modeling, Network Diffusion Process, and Bounded Confidence Models (BC Models) to define our model. We identify the agents involved in the communication process, the factors that affect the communication, different scenarios and the actions to be taken in these situations.…”

“…The model presented by (Bouanan, Zacharewicz, and Vallespir 2016) does not take into account the effects of varying the population size, the predefined threshold to adopt an opinion or the number of neighbors an individual has. To solve these limitations, in this paper, we include these parameters in the model.…”

“…As discussed in the introduction, based on the model presented by (Bouanan, Zacharewicz, and Vallespir 2016), we will use Cell-DEVS formalism (based on DEVS), Agent-Based Modeling, Network Diffusion Process, and Bounded Confidence Models (BC Models) to define a model to study the effect of different parameters in the human opinion of a product launch. Our goal is to provide a serious game that managers can use to understand the effects of human social interactions on others opinion.…”

“…Therefore, BC Models avoid the polarization by not generating the extremes. (Bouanan, Zacharewicz, and Vallespir 2016) used this combination of ABM and BC Models to study human behavior. They model the evolution of agent's opinion using Cell-DEVS formalism to validate the use of DEVS and Cell-DEVS in the study social influence using ABM.…”