Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)  2026, Vol. 60 Issue (6): 1261-1268    DOI: 10.3785/j.issn.1008-973X.2026.06.013
计算机技术     
融入人格特质的网络舆情风险预警方法
林浩1,2(),李雷孝1,2,*(),赵丽3
1. 内蒙古工业大学 智能科学与技术学院（网络空间安全学院），内蒙古 呼和浩特 010080
2. 内蒙古自治区北疆网络空间安全重点实验室，内蒙古 呼和浩特 010080
3. 天津理工大学 计算机科学与工程学院，天津 300384
Network public opinion risk warning method integrating personality trait
Hao LIN1,2(),Leixiao LI1,2,*(),Li ZHAO3
1. College of Intelligent Science and Technology (College of Cyberspace Security), Inner Mongolia University of Technology, Hohhot 010080, China
2. Inner Mongolia Key Laboratory of Beijiang Cyberspace Security, Hohhot 010080, China
3. School of Computer Science and Engineering, Tianjin University of Technology, Tianjin 300384, China
 全文: PDF(1757 KB)   HTML
摘要：

为了提高网络舆情风险预警方法的有效性和可解释性，以人格心理学和新社会分析模型为基础，提出融入人格特质信息的网络舆情预警指标. 结合事件主题、公众态度、情感倾向等主流指标，建立包含4个一级指标、18个二级指标的社交网络舆情预警指标体系. 采用熵权法确定各指标权重，利用TOPSIS综合评估舆情风险. 利用改进的灰狼算法优化的N-BEATS预测下一时刻的舆情风险. 为了验证该预警方法的可行性，采用所提方法预测分析4个真实事件. 因子分析结果表明，与情感倾向指标相比，提出的人格特质相关指标更重要. 预测结果表明，利用该预警方法能够拟合风险值序列，可以精确预测舆情风险.
关键词： 网络舆情;  舆情风险预警;  人格特质;  时间序列预测    
Abstract:

Online public opinion early warning indicator incorporating personality trait information was proposed based on personality psychology and the new social analysis model in order to enhance the effectiveness and interpretability of online public opinion risk early warning method. A social network public opinion early warning indicator system which comprised 4 first-level indicators and 18 second-level indicators was established by combining mainstream indicator such as event topics, public attitude and sentiment tendency. The entropy weight method was employed to determine the weight of each indicator, and TOPSIS was utilized for comprehensive evaluation of public opinion risk. An improved grey wolf optimizer-enhanced N-BEATS model was used to predict public opinion risk at the next moment. Four real-world events were analyzed using the proposed approach in order to verify the feasibility of the proposed early warning method. Factor analysis results show that personality trait-related indicator is more significant compared with sentiment tendency indicator. The prediction results demonstrate that the proposed early warning method can effectively fit risk value sequence and accurately predict public opinion risk.
Key words: network public opinion    public opinion risk warning    personality trait    time series prediction
收稿日期: 2025-07-07 出版日期: 2026-05-06
CLC:  G 353  
基金资助: 国家自然科学基金资助项目（62362055）；自治区首批“五大任务”关键技术研究专项资助项目（NMGWDRW2025-03）；内蒙古自治区重点研发与成果转化计划资助项目（2024SKYPT0012）；内蒙古自治区高等学校青年科技英才支持计划资助项目（NJYT22084）；内蒙古自然科学基金资助项目（2023MS06008）.
通讯作者: 李雷孝     E-mail: suzukaze_aoba@126.com;llxhappy@126.com
作者简介: 林浩（1995—），男，讲师，博士，从事数据挖掘、网络安全、人格计算的研究. orcid.org/0000-0001-9304-0279. E-mail：suzukaze_aoba@126.com
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章　　
林浩
李雷孝
赵丽

引用本文:

林浩,李雷孝,赵丽. 融入人格特质的网络舆情风险预警方法[J]. 浙江大学学报(工学版), 2026, 60(6): 1261-1268.

Hao LIN,Leixiao LI,Li ZHAO. Network public opinion risk warning method integrating personality trait. Journal of ZheJiang University (Engineering Science), 2026, 60(6): 1261-1268.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2026.06.013    或    https://www.zjujournals.com/eng/CN/Y2026/V60/I6/1261

图 1  网络舆情与新社会分析模型
图 2  大五人格特质与社交媒体行为之间的关系模型
图 3  提出的网络舆情风险预警方法
图 4  融入人格特质的网络舆情风险预警指标体系
预警等级风险评估值表述
Ⅰ级0~0.25网民关注度低，舆情传播速度慢，舆情影响局限在较小范围内，没有转化为行为舆论的可能
Ⅱ级0.25~0.50网民关注度中等，舆情传播速度中等，舆情影响局限在一定范围内，基本没有转化为行为舆论的可能
Ⅲ级0.50~0.75网民关注度高，舆情传播速度快，境外媒体可能关注，舆情影响范围很大，有转化为行为舆论的可能
Ⅳ级0.75~1.00网民关注度极高，境外媒体可能高度关注，舆情传播速度极快，舆情影响扩大到了整个社会，
有很大可能即将化为行为舆论
表 1  网络舆情风险预警等级
图 5  N-BEATS网络结构
参数取值范围
$ {t}_{\rm{SL}} $的长度系数$ {N}^{\text{predict}} $$ {N}^{\text{predict}}\in \left\{2,3,4,5,6,7\right\} $
前4层隐含层单元
数量$ {n}_{\rm{HU}} $		$ {n}_{\rm{HU}}\in \left[10，500\right]，{n}_{\rm{HU}}\in {\bf{N}}^{+} $
有2层隐含层单元
数量$ {n}_{\rm{TD}} $		$ {n}_{\rm{TD}}\in \left[1，20\right]，{n}_{\rm{TD}}\in {\bf{N}}^{+} $
N-BEATS中Stake
数量$ {n}_{\rm{SN}} $		$ {n}_{\rm{SN}}\in \left\{0,1,2,3,4,5,6,7,8\right\} $
Block类型$ {\alpha }_{\rm{BT}} $$ {\alpha }_{\rm{BT}}\in \left\{0,1,2\right\} $
Stake中Block的
数量$ {n}_{\rm{BPS}} $		$ {n}_{\rm{BPS}}\in \left[1，10\right]，{n}_{\rm{BPS}}\in {\bf{N}}^{+} $
权重$ {\alpha }_{\text{share}} $$ {\alpha }_{\text{share}}\in \left\{0,1\right\} $
批大小$ {n}_{\rm{BS}} $$ {n}_{\rm{BS}}\in \left[8，512\right]，{n}_{\rm{BS}}\in {\bf{N}}^{+} $
学习率$ {r}_{{\mathrm{l}}} $$ {r}_{\rm{l}}\in \left[0.000\;1,\;0.1\right] $
权重衰减系数$ {r}_{\rm{WD}} $$ {r}_{\rm{WD}}\in \left[0,1.0\right] $
一阶矩估计的指数
衰减率$ {r}_{\rm{EDR}} $		$ {r}_{\rm{EDR}}\in \left[0,1.0\right] $
二阶矩估计的指数
衰减率$ {{{r}^{\prime}}}_{\rm{EDR}} $		$ {{{r}^{\prime}}}_{\rm{EDR}}\in \left[0,1.0\right] $
表 2  需要优化的N-BEATS超参数
社交网络舆情案例舆情
类型		开始时间结束时间社交
媒体
延迟法定退休年龄改革政治2024-09-102024-09-15微博
三只羊美诚月饼事件社会2024-09-122024-09-27微博
上海大学生歌词歧视农民工社会2024-08-102024-08-18微博
《Mirror 2》游戏内容阉割文化2022-08-012022-12-31Steam
表 3  真实网络舆情案例的数据描述
图 6  利用熵权法求出的各案例对应的权重
社交网络舆情案例评论数转发数评论正面比例/%正面极性风险峰值预警等级
延迟法定退休年龄改革10485010798266.250.9730.680Ⅲ级
三只羊美诚月饼事件10790010478050.370.9740.328Ⅱ级
上海大学生歌词歧视农民工11512320739.690.9440.236Ⅰ级
《Mirror 2》游戏内容阉割7628—20.920.9160.153Ⅰ级
表 4  真实网络的舆情案例预警结果
图 7  利用DIGWO-N-BEATS拟合并预测各案例对应的风险值
1 王世文, 杨晨雁, 刘劲, 等 面向重大突发事件网络舆情的元数据框架研究[J]. 情报杂志, 2023, 42 (11): 105- 112
WANG Shiwen, YANG Chenyan, LIU Jin, et al Research on the metadata framework for network public opinion in major emergencies[J]. Journal of Intelligence, 2023, 42 (11): 105- 112
doi: 10.3969/j.issn.1002-1965.2023.11.016
2 MISLOVE A, LEHMANN S, AHN Y Y, et al Understanding the demographics of twitter users[J]. Proceedings of the International AAAI Conference on Web and Social Media, 2011, 5 (1): 554- 557
doi: 10.1609/icwsm.v5i1.14168
3 王宁, 赵胜洋, 单晓红 基于灰色系统理论的网络舆情预测与分级方法研究[J]. 情报理论与实践, 2019, 42 (2): 120- 126
WANG Ning, ZHAO Shengyang, SHAN Xiaohong Method of network public opinion prediction and grading based on grey system theory[J]. Information Studies: Theory and Application, 2019, 42 (2): 120- 126
doi: 10.16353/j.cnki.1000-7490.2019.02.020
4 朱国庆, 陈燕, 郎坤, 等 信息生态视角下森林火灾网络舆情风险预警研究[J]. 东北大学学报: 自然科学版, 2023, 44 (8): 1208- 1216
ZHU Guoqing, CHEN Yan, LANG Kun, et al Research on risk warning of network public opinion for forest fires from the perspective for information ecology[J]. Journal of Northeastern University: Natural Science, 2023, 44 (8): 1208- 1216
5 李绍攀, 赵飞, 周义棋, 等 基于微博数据的台风灾害舆情分析与灾害损失估计[J]. 清华大学学报: 自然科学版, 2022, 62 (1): 43- 51
LI Shaopan, ZHAO Fei, ZHOU Yiqi, et al Analysis of public opinion and disaster loss estimates from typhoons based on Microblog data[J]. Journal of Tsinghua University: Science and Technology, 2022, 62 (1): 43- 51
doi: 10.16511/j.cnki.qhdxxb.2021.26.031
6 ZHAO D, WU Y, ZHANG H A situation awareness approach for network security using the fusion model[J]. Mobile Information Systems, 2022, 2022 (1): 6214738
7 江志英, 李宇洋, 李佳桐, 等 基于层次分析的长短记忆网络(AHP-LSTM)的食品安全网络舆情预警模型[J]. 北京化工大学学报: 自然科学版, 2021, 48 (6): 98- 107
JIANG Zhiying, LI Yuyang, LI Jiatong, et al An early-warning model based on an analytic hierarchy process-long short-term memory network (AHP-LSTM) for food safety network public opinion[J]. Journal of Beijing University of Chemical Technology: Natural Science Edition, 2021, 48 (6): 98- 107
8 WEN Z, ZHANG L, WU Q, et al A network security situation awareness method based on GRU in big data environment[J]. International Journal of Pattern Recognition and Artificial Intelligence, 2023, 37: 2251018
doi: 10.1142/S0218001422510181
9 ROBERTS B W, WOOD D. Personality development in the context of the neo-socioanalytic model of personality [M]//Handbook of personality development. Mahwah: Lawrence Erlbaum Associates Publishers, 2006.
10 NYE C D, ROBERTS B W. A neo-socioanalytic model of personality development [M]//Work across the lifespan. London: Academic Press, 2019: 47-79.
11 LIN H, WANG C, SUN Y How big five personality traits influence information sharing on social media: a meta analysis[J]. PLoS One, 2024, 19 (6): e0303770
doi: 10.1371/journal.pone.0303770
12 PICAZO-VELA S, CHOU S Y, MELCHER A J, et al Why provide an online review? an extended theory of planned behavior and the role of Big-Five personality traits[J]. Computers in Human Behavior, 2010, 26 (4): 685- 696
doi: 10.1016/j.chb.2010.01.005
13 迪路阳, 钟寒, 施水才 网络舆情预警研究综述[J]. 数据分析与知识发现, 2023, 7 (8): 17- 29
DI Luyang, ZHONG Han, SHI Shuicai Review of early warning for online public opinion[J]. Data Analysis and Knowledge Discovery, 2023, 7 (8): 17- 29
14 李振鹏, 陈碧珍, 罗静宇 基于文本挖掘的网络舆情分类研究[J]. 系统科学与数学, 2020, 40 (5): 813- 826
LI Zhenpeng, CHEN Bizhen, LUO Jingyu Research on network public opinion classification based on text mining[J]. Journal of Systems Science and Mathematical Sciences, 2020, 40 (5): 813- 826
15 LIN H, WANG C, HAO Q A novel personality detection method based on high-dimensional psycholinguistic features and improved distributed Gray Wolf Optimizer for feature selection[J]. Information Processing and Management, 2023, 60 (2): 103217
doi: 10.1016/j.ipm.2022.103217
16 LUO Z, TIAN J, ZENG J, et al Flood risk evaluation of the coastal city by the EWM-TOPSIS and machine learning hybrid method[J]. International Journal of Disaster Risk Reduction, 2024, 106: 104435
doi: 10.1016/j.ijdrr.2024.104435
17 JI J, WANG D Evaluation analysis and strategy selection in urban flood resilience based on EWM-TOPSIS method and graph model[J]. Journal of Cleaner Production, 2023, 425: 138955
doi: 10.1016/j.jclepro.2023.138955
18 ORESHKIN B N, CARPOV D, CHAPADOS N, et al. N-BEATS: neural basis expansion analysis for interpretable time series forecasting [EB/OL]. [2025-06-15]. https://arxiv.org/abs/1905.10437.
19 LIN H, WANG C DIGWO-N-BEATS: an evolutionary time series prediction method for situation prediction[J]. Information Sciences, 2024, 664: 120316
doi: 10.1016/j.ins.2024.120316
20 崔鹏, 张巍, 何毅, 等 突发公共事件网络舆情演化及政府应对能力研究[J]. 现代情报, 2018, 38 (2): 75- 83
CUI Peng, ZHANG Wei, HE Yi, et al Dynamic evolution research on the government’s response capability to the public opinions in the context of public emergencies[J]. Journal of Modern Information, 2018, 38 (2): 75- 83
doi: 10.3969/j.issn.1008-0821.2018.02.012
[1] 刘瑄昀,闫莹,於志勇,黄昉菀. 基于无标度网络的类脑储备池拓扑设计[J]. 浙江大学学报(工学版), 2025, 59(7): 1385-1393.
[2] 疏阳,孙轶琳,梅振宇,张逸敏,黄毅方. 基于多源数据的个体活动序列多步预测[J]. 浙江大学学报(工学版), 2025, 59(11): 2317-2325.
[3] 马泽超,刘小明,夏汗青,王伟强,王久增,申海涛. 基于图神经网络的路面病害态势预测方法[J]. 浙江大学学报(工学版), 2024, 58(12): 2596-2608.
[4] 王晨霖,杨洁,居文军,顾复,陈芨熙,纪杨建. 基于智能家电的短期电力负荷预测与削峰填谷优化[J]. 浙江大学学报(工学版), 2020, 54(7): 1418-1424.
[5] 汪子龙,王柱,於志文,郭斌,周兴社. 多源数据跨国人口迁移预测[J]. 浙江大学学报(工学版), 2019, 53(9): 1759-1767.