突发自然灾害下短视频网络舆情的互动演化与治理研究

Study on the interaction evolution and governance of short video online public opinion under sudden natural disasters

摘要

本文基于短视频的多模态数据,解析突发自然灾害情境下网络舆情的互动演化过程与特征,探索影响短视频网络舆情互动指数的关键要素。采用量化分析范式,应用机器学习模型极端梯度提升(eXtreme gradient boosting,XGBoost)算法与沙普利可解释性分析(SHapley Additive exPlanations,SHAP)方法测度短视频舆情特征变量。结果表明,短视频舆情演化具有阶段性特征,弹幕和评论数在突发事件发生后的短期内迅速上升,并随时间推移逐渐回落。视频播放数、用户满意度指数、平均弹幕和评论字数等特征对互动指数具有显著正向影响,满意度指数的影响最为稳定。视频作者投稿数、作者影响力对互动指数的影响呈现负相关关系,积极情绪倾向对互动指数的影响呈现非线性特征。基于研究结果提出舆情预警机制、内容引导策略,以及多主体协同治理模式的多种舆情治理优化策略,以提升短视频舆情应对的科学性和有效性,为突发自然灾害背景下的网络舆情管理提供理论支撑和实践参考。

Abstract

This paper analyzes the interactional evolution and characteristics of online public opinion in the context of sudden natural disasters based on multimodal short video data, explores the key factors influencing the short video public opinion interaction index. A quantitative analytical paradigm is adopted, employing the XGBoost (eXtreme gradient boosting) machine learning model and SHAP (SHapley Additive exPlanations) interpretability analysis to evaluate variables of short video public opinion. The results reveal that short video public opinion exhibits a staged evolutionary pattern, the number of bullet comments and text comments surges shortly after the occurrence of a sudden event and gradually declines over time. Features such as video views, user satisfaction index, and the average number of characters in bullet comments and text comments have a significant positive effect on the interaction index, with the user satisfaction index being the most stable contributor. In contrast, the number of video author submissions and the author influence show a negative correlation with the interaction index. Moreover, the impact of positive emotional tendencies on interaction index exhibits a non-linear characteristic. Based on the results, several strategies to optimize public opinion governance are proposed, including a public opinion early warning mechanism, content guidance strategies, and a multi-agent collaborative governance model, to enhance the scientific and effective response to short video public opinion, and provide theoretical and practical support for managing online public opinion in the context of sudden natural disasters.

关键词

自然灾害 / 短视频 / 网络舆情 / 机器学习 / XGBoost / SHAP

Key words

natural disasters / short videos / network public opinion / machine learning / XGBoost / SHAP

参考文献

[1] 周江萍, 黄国平. 自然灾害应急物流能力关键影响因素及其结构识别[J]. 自然灾害学报, 2025, 34(2): 1-12.

[2] Zhou Jiangping, Huang Guoping. Identification of key influencing factors and their structure of natural disaster emergency logistics capacity[J]. Journal of Natural Disasters, 2025, 34(2): 1-12. (in Chinese)

[3] 赖茂生, 李爱新, 梅培培. 信息生命周期管理理论与政府信息资源管理创新研究[J]. 图书情报工作, 2014, 58(6): 6-11,41.

[4] Lai Maosheng, Li Aixin, Mei Peipei. Study on information life cycle management theory and government information resources management innovation[J]. Library and Information Service, 2014, 58(6): 6-11,41. (in Chinese)

[5] 洪亮, 石立艳, 李明. 基于系统动力学的多主体回应网络舆情影响因素研究[J]. 情报科学, 2017, 35(1): 133-138.

[6] Hong Liang, Shi Liyan, Li Ming. Research on multi-agent response factors of network public opinion based on system dynamics[J]. Information Science, 2017, 35(1): 133-138. (in Chinese)

[7] Steinfeld N, Samuel-Azran T, Lev-On A. User comments and public opinion: findings from an eye-tracking experiment[J]. Computers in Human Behavior, 2016, 61: 63-72.

[8] Geng Lixiao, Yang Shuran, Wang Ke, et al. Modeling public opinion dissemination in a multilayer network with SEIR model based on real social networks[J]. Engineering Applications of Artificial Intelligence, 2023, 125: 106719.

[9] 张桂蓉, 董志香, 夏霆. 突发公共卫生事件网络谣言网格化预警模型研究[J]. 中国管理科学, 2024, 32(2): 298-306.

[10] Zhang Guirong, Dong Zhixiang, Xia Ting. Study on the grid-based early warning model for online rumors of public health emergencies[J]. Chinese Journal of Management Science, 2024, 32(2): 298-306. (in Chinese)

[11] 李伟卿, 汪文涛, 黄炜, 等 . 弹幕视频网站用户持续使用行为的影响因素及其可解释性分析: 基于感知价值的视角[J]. 现代情报, 2023, 43(12): 63-72.

[12] Li Weiqing, Wang Wentao, Huang Wei, et al. Study on the influencing factors and its explainability of users’ continuous use behavior of bullet screen video website: perspective based on perceived value[J]. Journal of Modern Information, 2023, 43(12): 63-72. (in Chinese)

[13] 王文韬, 宋天骁, 唐思捷, 等 . 短视频用户间歇性中辍行为反复机理及优化研究[J]. 现代情报, 2023, 43(12): 51-62.

[14] Wang Wentao, Song Tianxiao, Tang Sijie, et al. Research on the recurrent mechanism and optimization of intermittent dropout behavior of short video users[J]. Modern Information, 2023, 43(12): 51-62. (in Chinese)

[15] 黄鑫昊, 冯馨瑶. 大学生移动短视频主动发布意愿的动力机制研究[J]. 情报科学, 2021, 39(11): 83-89, 95.

[16] Huang Xinhao, Feng Xinyao. The dynamic mechanism of college students’ willingness to actively publish mobile short video[J]. Information Science, 2021, 39(11): 83-89, 95. (in Chinese)

[17] 孙平, 邵帅, 石佳云, 等 . 基于扎根理论的短视频抖音用户出游行为形成机理研究[J]. 管理学报, 2020, 17(12): 1823-1830.

[18] Sun Ping, Shao Shuai, Shi Jiayun, et al. Formation mechanism of TikTok users’ travel behavior: an exploration study based on grounded theory[J]. Chinese Journal of Management, 2020, 17(12): 1823-1830. (in Chinese)

[19] 王伟军, 叶建梅, 周锐, 等 . UGC短视频用户持续贡献行为影响因素及作用机理研究[J]. 图书与情报, 2022(5): 31-40.

[20] Wang Weijun, Ye Jianmei, Zhou Rui, et al. Research on the factors influencing the continuous contribution behavior of UGC short video users and the mechanism of their role[J]. Library & Information, 2022(5): 31-40. (in Chinese)

[21] Vosoughi S, Roy D, Aral S. The spread of true and false news online[J]. Science, 2018, 359(6380): 1146-1151.

[22] 兰月新, 李奇峰, 何巍, 等 . 兼听则明:基于群体情感熵的网络舆情认知风险防范机制[J]. 现代情报, 2025, 45(12): 103-117.

[23] Lan Yuexin, Li Qifeng, He Wei, et al. Listening to both sides leads to enlightenment: a cognitive risk prevention mechanism for online public opinion based on group emotional entropy[J]. Journal of Modern Information, 2025, 45(12): 103-117. (in Chinese)

[24] 岳丽欣, 刘自强, 朱承宁. 突发自然灾害事件短视频用户情感时空演化分析方法研究[J]. 图书情报工作, 2024, 68(24): 77-91.

[25] Yue Lixin, Liu Ziqiang, Zhu Chengning. Research on the analysis method of user emotional spatiotemporal evolution in short video related to sudden natural disaster events[J]. Library and Information Service, 2024, 68(24): 77-91. (in Chinese)

[26] 叶凤云, 白佼璐, 李君君. 短视频用户偶遇信息分享行为影响因素研究[J]. 情报理论与实践, 2023, 46(10): 121-129.

[27] Ye Fengyun, Bai Jiaolu, Li Junjun. Research on influencing factors of short video users’ encounter information sharing behavior[J]. Information Studies (Theory & Application), 2023, 46(10): 121-129. (in Chinese)

[28] 赵月华, 盛予欣. 短视频平台用户交互行为的差异性[J]. 图书馆论坛, 2024, 44(2): 114-122.

[29] Zhao Yuehua, Sheng Yuxin. Differences in user interaction behaviour on short video platforms[J]. Library Tribune, 2024, 44(2): 114-122. (in Chinese)

[30] 梁循, 王涵玉, 张森森, 等 . 人工智能生成内容对舆情生态的影响、挑战和治理[J]. 管理世界, 2025, 41(6): 129-157.

[31] Liang Xun, Wang Hanyu, Zhang Sensen, et al. Impacts, challenges, and governance of artificial intelligence generated content on the public opinion ecosystem[J]. Journal of Management World, 2025, 41(6): 129-157. (in Chinese)

[32] Hoiles W, Aprem A, Krishnamurthy V. Engagement and popularity dynamics of YouTube videos and sensitivity to meta-data[J]. IEEE Transactions on Knowledge and Data Engineering, 2017, 29(7): 1426-1437.

[33] Feroz Khan G, Vong S. Virality over YouTube: an empirical analysis[J]. Internet Research, 2014, 24(5): 629-647.

[34] Zhang Xiaodan, Zhao Zhanbo, Wang Kui. The effects of live comments and advertisements on social media engagement: application to short-form online video[J]. Journal of Research in Interactive Marketing, 2024, 18(3): 485-505.

[35] 魏宏程, 朱恒民, 魏静, 等 . 基于短视频网络的互联网舆情演化研究[J]. 数据分析与知识发现, 2024, 8(5): 113-126.

[36] Wei Hongcheng, Zhu Hengmin, Wei Jing, et al. Analyzing the evolution of Internet public opinion based on short-video network[J]. Data Analysis and Knowledge Discovery, 2024, 8(5): 113-126. (in Chinese)

[37] 黄方楠, 李明德. 短视频舆情的非理性表达与互动分析[J]. 情报杂志, 2022, 41(3): 152-158,113.

[38] Huang Fangnan, Li Mingde. Irrational expression and interaction of online public opinion in short video[J]. Journal of Intelligence, 2022, 41(3): 152-158,113. (in Chinese)

[39] 刘耀辉, 刘婉婷, 张文焯, 等 . 基于新浪微博数据的云南漾濞6.4级地震舆情时空特征及情感分析[J]. 自然灾害学报, 2022, 31(1): 168-178.

[40] Liu Yaohui, Liu Wanting, Zhang Wenzhuo, et al. Spatiotemporal characteristics of public opinion and emotion analysis of M_S6.4 Yunnan Yangbi earthquake based on Sina Weibo data [J]. Journal of Natural Disasters, 2022, 31(1): 168-178. (in Chinese)

基金资助

北京市社会科学基金项目(19GLB040)