Fusing Dynamics Equation: A Social Opinions Prediction Algorithm with LLM-based Agents

Junchi Yao,Hongjie Zhang,Jie Ou,Dingyi Zuo,Zheng Yang,Zhicheng Dong

2024-09-27

Abstract:In the context where social media is increasingly becoming a significant platform for social movements and the formation of public opinion, accurately simulating and predicting the dynamics of user opinions is of great importance for understanding social phenomena, policy making, and guiding public opinion. However, existing simulation methods face challenges in capturing the complexity and dynamics of user behavior. Addressing this issue, this paper proposes an innovative simulation method for the dynamics of social media user opinions, the FDE-LLM algorithm, which incorporates opinion dynamics and epidemic model. This effectively constrains the actions and opinion evolution process of large language models (LLM), making them more aligned with the real cyber world. In particular, the FDE-LLM categorizes users into opinion leaders and followers. Opinion leaders are based on LLM role-playing and are constrained by the CA model, while opinion followers are integrated into a dynamic system that combines the CA model with the SIR model. This innovative design significantly improves the accuracy and efficiency of the simulation. Experiments were conducted on four real Weibo datasets and validated using the open-source model ChatGLM. The results show that, compared to traditional agent-based modeling (ABM) opinion dynamics algorithms and LLM-based opinion diffusion algorithms, our FDE-LLM algorithm demonstrates higher accuracy and interpretability.

Social and Information Networks,Computers and Society

What problem does this paper attempt to address?

The paper attempts to address the problem of how to accurately simulate and predict the dynamic changes in user opinions in the context of social media becoming an increasingly important platform for social movements and public opinion formation. Existing simulation methods face challenges in capturing the complexity and dynamics of user behavior. Therefore, this paper proposes an innovative simulation method—the FDE-LLM algorithm, which aims to improve the accuracy and efficiency of simulations by combining opinion dynamics models and epidemic models.

Specifically, the FDE-LLM algorithm divides users into opinion leaders and followers. Opinion leaders role-play based on large language models (LLM) and are constrained by the cellular automaton (CA) model; while opinion followers are integrated into a dynamic system that combines the CA model and the SIR model. This design makes the simulation results closer to the real network world, improving the accuracy and interpretability of the simulations.

By conducting experiments on 4 real Weibo datasets and comparing with traditional agent-based modeling (ABM) opinion dynamics algorithms and LLM-based opinion propagation algorithms, the results show that the FDE-LLM algorithm performs excellently in terms of dynamic time warping (DTW) distance and Pearson correlation coefficient metrics.