Online opinion dynamics model considering derivative topics based on complex network
- 30 June 2025
- journal article
- research article
- Published by Emerald Publishing in Information Technology & People
- Vol. 39 (3) , 1341-1364
Abstract
Purpose: In social networks, derivative topics derived from original topics are becoming increasingly general, and the interaction between derivative topics and original topics makes public opinions increasingly complex and unstable. This paper aims to study the evolution of online opinions under the interaction between original and derivative topics. Design/methodology/approach: This paper establishes an online opinion dynamics model based on a two-layer complex network, consisting of original and derivative topic layers. In the two-layer network, the interaction threshold between netizens and the external environment is set dynamically. Secondly, by setting herd mentality, response subject, and media parameters, the internal psychological factors and external environmental factors are involved in the model. Finally, simulation experiments are conducted using actual online social network data to examine the effects of topic adherence, relative clout of topics, and external environmental information on the opinions related to the original and derivative topics. Findings: Topic relationships are crucial in the interaction between original and derivative topics. Derivative topics that are highly related to the original topic will have a greater impact on each other’s dissemination. Those with high relative clout can influence the original topic more rapidly. Therefore, when encountering derivative topics that are highly related to the original topic or have high relative clout, relevant departments should give greater and more timely attention. In addition, due to topic relationships, governing a single topic always leads to an increase in negative opinions on another topic, which may turn low-clout topics into high-clout ones. Thus, the response subject can actively explore the commonalities between derivative and original topics and intervene in common issues to enhance effectiveness. Research limitations/implications: First, our model did not consider forgetting mechanisms and node decay effects which may affect the long-term dissemination of information and the intensity of interaction between nodes. Second, we treat the response subject’s credibility and information quality as fixed parameters. Third, we employ a rough estimation method to assess topic adherence. Practical implications: Due to topic relationships, the governance of a single topic always leads to an increase in negative opinions on another topic, which may turn low-clout topics into high-clout topics. Thus, the response subject can actively explore the commonalities between the derivative and original topics and intervene in common issues to be more effective. Originality/value: This paper explores the interaction scenario between the original and derivative topics, providing valuable insights for more complex public opinion governance issues.Keywords
This publication has 45 references indexed in Scilit:
- A model for early rumor detection base on topic-derived domain compensation and multi-user associationExpert Systems with Applications, 2024
- A game approach to multi-dimensional opinion dynamics in social networks with stubborn strategist agentsEuropean Journal of Control, 2024
- Analysis of opinion evolution based on non-Bayesian social learningApplied Mathematics and Computation, 2023
- Opinion dynamics on social networksActa Mathematica Scientia, 2022
- The influence of communication structure on opinion dynamics in social networks with multiple true statesApplied Mathematics and Computation, 2021
- HK–SEIR model of public opinion evolution based on communication factorsEngineering Applications of Artificial Intelligence, 2021
- Temporal and spatial evolution of online public sentiment on emergenciesInformation Processing & Management, 2020
- Prediction of Individual’s Character in Social Media Using Contextual Semantic Sentiment AnalysisMobile Networks and Applications, 2019
- Sentiment analysis of Chinese micro-blog text based on extended sentiment dictionaryFuture Generation Computer Systems, 2018
- The fusion process of interval opinions based on the dynamic bounded confidenceInformation Fusion, 2016