突发事件网络的传播规律及特征 ——以知网事件为例
Network propagation law and characteristics of emergencies ——take the CNKI incident as an example
投稿时间: 2022/12/20 0:00:00
DOI:
中文关键词: 突发事件;意见领袖;社会网络分析;LDA主题提取;情感分析
英文关键词: emergency; opinion leaders; social network analysis; LDA topic extraction; sentiment analysis
基金项目: :北京市自然科学基金(9202018);中国传媒大学中央高校基本科研业务费专项资金资助(CUC21GZ005);中国广播电视协会学术研究项目(2020ZGLH015);国家自然科学基金(11901542)
姓名 单位
杨妺 中国传媒大学数据科学与智能媒体学院
王妍 中国传媒大学数据科学与智能媒体学院
王传彪 中国传媒大学数据科学与智能媒体学院
李昊展 中国传媒大学数据科学与智能媒体学院
刘铭 中国传媒大学数据科学与智能媒体学院
点击数:614 下载数:503
中文摘要:

以“知网事件”为例,构建微博转发网络,运用社会网络分析方法,从网络全局特征、网络结构特征、节点影响力三个角度,分析该事件传播规律,识别核心节点;采集意见领袖发布的文本信息,运用LDA主题提取以及SnowNLP情感分析,分析意见领袖的观点倾向。实例分析表明:此次事件参与传播的人数多、范围广,但个体化强,互相联系松散,意见领袖在转发网络中具有较大的出度、度中心性和接近中心性,以及较小的限制度,属于4-shell层。意见领袖发布内容均持中立偏积极的态度,主题为事件阐述和经营状况。

英文摘要:

Taking "CNKI" as an example, constructing a Weibo forwarding network, using social network analysis methods, from three perspectives of network global characteristics, network structure characteristics, and node influence,we analyzed the propagation law of the event, and identified the core Nodes. Then we collected text information released by opinion leaders, and used LDA topic extraction and SnowNLP sentiment analysis to analyze opinion leaders' opinion tendencies.The case analysis shows that the number of people involved in the communication of this event is large and the scope is wide, but the individuality is strong and the relationship is loose. The opinion leaders have a large out-degree, degree centrality and closeness centrality in the forwarding network, and belong to the 4-shell layer. The content released by opinion leaders has a neutral and positive attitude, and the themes are event descriptions and business conditions.

参考文献: