Abstract:This paper investigates how interacting agents arrive to a consensus or a polarized state. We study the opinion formation process under the effect of a global steering mechanism (GSM), which aggregates the opinion-driven stochastic agent states at the network level and feeds back to them a form of global information. We also propose a new two-layer agent-based opinion formation model, called GSM-DeGroot, that captures the coupled dynamics between agent-to-agent local interactions and the GSM's steering effect. This way, agents are subject to the effects of a DeGroot-like local opinion propagation, as well as to a wide variety of possible aggregated information that can affect their opinions, such as trending news feeds, press coverage, polls, elections, etc. Contrary to the standard DeGroot model, our model allows polarization to emerge by letting agents react to the global information in a stubborn differential way. Moreover, the introduced stochastic agent states produce event stream dynamics that can fit to real event data. We explore numerically the model dynamics to find regimes of qualitatively different behavior. We also challenge our model by fitting it to the dynamics of real topics that attracted the public attention and were recorded on Twitter. Our experiments show that the proposed model holds explanatory power, as it evidently captures real opinion formation dynamics via a relatively small set of interpretable parameters.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to study how interacting agents reach consensus or form a polarized state when introducing the Global Steering Mechanism (GSM). Specifically, the paper explores the process of individual opinion formation and its dynamic characteristics in the context of modern communication technologies, and proposes a new two - layer agent - based opinion - formation model - the GSM - DeGroot model. This model aims to capture the coupling dynamics between local interactions and global information guidance, thereby explaining the complex behavior of opinion formation.
### Specific description of the problem
1. **Dynamics of opinion formation and polarization**:
- Existing models, such as the classic DeGroot model, can describe the process of opinion convergence, but cannot explain the phenomena of opinion diversity and polarization. Therefore, this paper attempts to fill this gap by introducing the global steering mechanism.
2. **Interaction between local and global information**:
- In the real world, individuals are influenced not only by local social networks, but also by mass media and other global information sources. Existing models usually ignore the role of this global information, and this paper attempts to simulate this influence through GSM.
3. **Event - driven opinion expression**:
- This paper also considers the behavior of individuals expressing opinions on public platforms (such as posting, participating in protests, etc.), which can be regarded as opinion - dependent random events. By introducing the Event Generation Mechanism (EGM), the paper attempts to more comprehensively describe the behavioral characteristics of individuals in the opinion - formation process.
### Main contributions of the model
- **GSM - DeGroot model**: Combines the local Opinion Propagation Mechanism (OPM) and the global Steering Mechanism (GSM), and introduces the Event Generation Mechanism (EGM). This model allows individuals to have different reactions to global information, thereby explaining the phenomena of opinion diversity, polarization, and radicalization.
- **Mathematical properties and numerical simulations**: Through strict mathematical derivations and extensive numerical simulations, the paper shows the model behavior in different parameter regions and verifies the explanatory and fitting capabilities of the model.
### Formula summary
1. **State update formula**:
\[
S_{i,t} \sim \text{Bernoulli}\left(\frac{1}{1 + \exp(-\lambda X_{i,t})}\right)
\]
where \( S_{i,t} \) represents the random state of individual \( i \) at time \( t \), \( X_{i,t} \) is its opinion value, and \( \lambda \) is the sensitivity parameter.
2. **Opinion update formula**:
\[
X_{i,t + 1} = \beta_i g(S_t) + \sum_{j = 1}^N w_{ji} X_{j,t}
\]
where \( \beta_i \) represents the fixed reaction mode of individual \( i \) to global information, \( g(S_t) \) is the summary information of the global steering mechanism, and \( w_{ji} \) is the influence weight of individual \( j \) on individual \( i \).
3. **Average opinion change formula**:
\[
X_{t + 1} = (2\beta - 1)\gamma \tilde{S}_t + X_t
\]
where \( \beta \) is the proportion of individuals with a positive reaction to global information, \( \gamma \) is the scaling parameter of the global steering mechanism, and \( \tilde{S}_t \) is the proportion of global information summary.
Through these formulas and models, the paper successfully describes the complex dynamics of the interaction between local and global information in the opinion - formation process, and provides a new perspective for understanding the evolution of public opinion in modern society.