| 于凯,杨富义.社会安全事件网络舆情多属性演化分析与知识图谱构建[J].情报工程,2022,8(4):014-030 |
| 社会安全事件网络舆情多属性演化分析与知识图谱构建 |
| Multi-Attribute Evolution Analysis and Knowledge Graph Construction of Network Public Opinion on Social Security Events |
| |
| DOI：10.3772/j.issn.2095-915X.2022.04.002 |
| 中文关键词: 社会安全事件；网络舆情；知识图谱；主题挖掘；情感演化 |
| 英文关键词: Social security events; network public opinion; knowledge graph; topic mining; emotional evolution |
| 基金项目:新疆维吾尔自治区自然科学基金项目“基于多层网络模型的信息传播源头定位研究”(2019D01A22)；新疆维吾尔自治区社科基金项目“智能时代新疆重大舆情和突发事件舆论治理与引导机制研究”(21BTQ162)。 |
| 作者 | 单位 | | 于凯 | 1. 新疆财经大学公共管理学院 乌鲁木齐 830012； | | 杨富义 | 2. 新疆财经大学信息管理学院 乌鲁木齐 830012 |
|
| 摘要点击次数: 737 |
| 全文下载次数: 600 |
| 中文摘要: |
| [ 目的 / 意义 ] 挖掘社会安全事件网络舆情的传播规律与多属性演化特征，为网络舆情事件的应对和治理提供参考。[ 方法 / 过程 ] 以“货拉拉女乘客跳车事件”为例，从舆情主题、参与主体和网民情绪三个角度进行社会安全事件网络舆情多属性演化分析，探究舆情各阶段网民的关注焦点和情感动向；其次基于事理逻辑构建舆情事件知识图谱，以明确舆情事件间的关联关系和演化规律。[ 结果 / 结论 ] 不同阶段的舆情主题、关键主体与情感特征具有明显的区别。网络舆情多属性演化分析有助于舆情监管部门迅速把握事件的多属性演化动态，而舆情事件知识图谱可以厘清事件传播路径、定位事件关键治理节点并还原舆情全貌。[ 局限 ] 后续研究将考虑细粒度情感分析，同时结合事件及关系抽取模型构建舆情事件知识图谱，扩展事件类型和数据规模。 |
| 英文摘要: |
| [Objective/Significance] Excavate the propagation law and multi-attribute evolution characteristics of network public opinion on social security events, so as to provide reference for the response and governance of network public opinion events. [Methods/Processes] Taking the “LaLaMove Female Passenger Jumping Incident” as an example, this paper analyzes the multiattribute evolution of social security events’ network public opinion from the perspectives of public opinion theme, participants and netizens’ emotions, and explores the focus and emotional trends of netizens at all stages of public opinion; Secondly, the knowledge graph of public opinion events is constructed based on the logic of affairs to clarify the correlation and evolution law between public opinion events [Results/Conclusions] There are obvious differences in public opinion themes, key subjects and emotional characteristics in different stages. The multi-attribute evolution analysis of network public opinion help public opinion regulators quickly grasp the multi-attribute evolution dynamics of events, and the knowledge graph of public opinion events can clarify the propagation path of events, locate the key governance nodes of events, and restore the overall picture of public opinion. [Limitations] The follow-up research will consider fine-grained emotion analysis, and combine the event and relationship extraction model to build the knowledge graph of public opinion events, so as to expand the event type and data scale. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
