Social opinions prediction utilizes fusing dynamics equation with LLM-based agents

Abstract In the context where social media emerges as a pivotal platform for social movements and shaping public opinion, accurately simulating and predicting the dynamics of user opinions is of significant importance. Such insights are vital for understanding social phenomena, informing policy deci...

Full description

Saved in:
Bibliographic Details
Main Authors: Junchi Yao, Hongjie Zhang, Jie Ou, Dingyi Zuo, Zheng Yang, Zhicheng Dong
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-025-99704-3
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850206230830645248
author Junchi Yao
Hongjie Zhang
Jie Ou
Dingyi Zuo
Zheng Yang
Zhicheng Dong
author_facet Junchi Yao
Hongjie Zhang
Jie Ou
Dingyi Zuo
Zheng Yang
Zhicheng Dong
author_sort Junchi Yao
collection DOAJ
description Abstract In the context where social media emerges as a pivotal platform for social movements and shaping public opinion, accurately simulating and predicting the dynamics of user opinions is of significant importance. Such insights are vital for understanding social phenomena, informing policy decisions, and guiding public opinion. Unfortunately, traditional algorithms based on idealized models and disregarding social data often fail to capture the complexity and nuance of real-world social interactions. This study proposes the Fusing Dynamics Equation-Large Language Model (FDE-LLM) algorithm. This innovative approach aligns the actions and evolution of opinions in Large Language Models (LLMs) with the real-world data on social networks. The FDE-LLM divides users into two roles: opinion leaders and followers. Opinion leaders use LLM for role-playing and employ Cellular Automata(CA) to constrain opinion changes. In contrast, opinion followers are integrated into a dynamic system that combines the CA model with the Susceptible-Infectious-Recovered (SIR) model. This innovative design significantly improves the accuracy of the simulation. Our experiments utilized four real-world datasets from Weibo. The result demonstrates that the FDE-LLM significantly outperforms traditional Agent-Based Modeling (ABM) algorithms and LLM-based algorithms. Additionally, our algorithm accurately simulates the decay and recovery of opinions over time, underscoring LLMs potential to revolutionize the understanding of social media dynamics.
format Article
id doaj-art-ade9e048f4a8449c8f9950f8c051d27c
institution OA Journals
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-ade9e048f4a8449c8f9950f8c051d27c2025-08-20T02:10:53ZengNature PortfolioScientific Reports2045-23222025-05-0115111410.1038/s41598-025-99704-3Social opinions prediction utilizes fusing dynamics equation with LLM-based agentsJunchi Yao0Hongjie Zhang1Jie Ou2Dingyi Zuo3Zheng Yang4Zhicheng Dong5College of Computer Science, Sichuan Normal UniversityCollege of Computer Science, Sichuan Normal UniversitySchool of Information and Software Engineering, University of Electronic Science and Technology of ChinaCollege of Computer Science, Sichuan Normal UniversityFujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal UniversityCollege of Information Science Technology, Tibet UniversityAbstract In the context where social media emerges as a pivotal platform for social movements and shaping public opinion, accurately simulating and predicting the dynamics of user opinions is of significant importance. Such insights are vital for understanding social phenomena, informing policy decisions, and guiding public opinion. Unfortunately, traditional algorithms based on idealized models and disregarding social data often fail to capture the complexity and nuance of real-world social interactions. This study proposes the Fusing Dynamics Equation-Large Language Model (FDE-LLM) algorithm. This innovative approach aligns the actions and evolution of opinions in Large Language Models (LLMs) with the real-world data on social networks. The FDE-LLM divides users into two roles: opinion leaders and followers. Opinion leaders use LLM for role-playing and employ Cellular Automata(CA) to constrain opinion changes. In contrast, opinion followers are integrated into a dynamic system that combines the CA model with the Susceptible-Infectious-Recovered (SIR) model. This innovative design significantly improves the accuracy of the simulation. Our experiments utilized four real-world datasets from Weibo. The result demonstrates that the FDE-LLM significantly outperforms traditional Agent-Based Modeling (ABM) algorithms and LLM-based algorithms. Additionally, our algorithm accurately simulates the decay and recovery of opinions over time, underscoring LLMs potential to revolutionize the understanding of social media dynamics.https://doi.org/10.1038/s41598-025-99704-3
spellingShingle Junchi Yao
Hongjie Zhang
Jie Ou
Dingyi Zuo
Zheng Yang
Zhicheng Dong
Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
Scientific Reports
title Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
title_full Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
title_fullStr Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
title_full_unstemmed Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
title_short Social opinions prediction utilizes fusing dynamics equation with LLM-based agents
title_sort social opinions prediction utilizes fusing dynamics equation with llm based agents
url https://doi.org/10.1038/s41598-025-99704-3
work_keys_str_mv AT junchiyao socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents
AT hongjiezhang socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents
AT jieou socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents
AT dingyizuo socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents
AT zhengyang socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents
AT zhichengdong socialopinionspredictionutilizesfusingdynamicsequationwithllmbasedagents