摘要
大数据环境下,为了实现对突发产品安全事件的网络舆情线索进行全面监测和快速通报,实现对海量舆情信息的传播规律和更有针对性地制定事件干预决策,提出了一种前置舆情分析和预警模型。该模型通过集合产品质量安全事件相关的网络舆情案例信息,基于知识图谱理论的领域本体描述拆解案例信息的故障伤害安全严重程度,利用文本聚类技术判断故障发生是否为批次性及批量性、利用互联网传播属性数据评估该事件的媒介权重、传播广度和受众深度,形成五维雷达图预警,在快速预警突发事件的同时直观展现舆情演进。
In the context of big data, in order to realize comprehensive monitoring and rapid notification of network public opinion clues for product safety incidents and the transmission law of huge amounts of public opinion information and to make targeted incident intervention decisions, this paper puts forth a model of pre-set public opinion analysis and early warning. This model can gather public opinion case information related to product quality safety incidents on the Internet, and uses the domain ontology based on the knowledge map theory to describe and disassemble the seriousness of injury of case information, then judges whether the failure occurs in batch. It uses the clustering technology to assess the medium weight, spreading breadth and audience depth, and applies the Internet transmission property data to form the five-dimensional radar map for early warning and intuitively displays the progress of public opinion along with rapid early warning of emergencies.
作者
姜肇财
孙宁
宋黎
JIANG Zhao-cai;SU N Ning;SO NG Li(China National Institute of Standardization)
出处
《标准科学》
2020年第7期44-49,共6页
Standard Science
基金
国家市场监督管理总局科技计划项目“基于缺陷汽车产品故障表现的互联网影响力评价研究”(项目编号:2017QK098)
中国标准化研究院院长基金项目“基于电商评价信息的典型消费品缺陷线索安全评价方法研究”(项目编号:282020Y-7511)资助。
关键词
产品安全事件
网络舆情
快速预警
product safety incidents
network public opinion
early warming
