摘要
为了提高新闻舆情发展判断的准确性,基于知识图谱的社会网络热点新闻舆情研究与判断方法成为了最新的研究课题,通过语料库标注、字符编码和时间片处理,对网络热点新闻语料库进行预处理,并利用处理后的语料库信息构建知识图谱来提高新闻要素分析的准确性.针对传统舆情系统的缺陷,从知识图谱的角度出发,运用文本向量化的方法来提高发现有价值信息的准确率.实验结果表明,该方法对舆情演变程度、传播广度和新闻热度的判断有较好的准确性及较小的误差.
In order to improve the accuracy of the judgment of news and public opinion development judgment,the research and judgment method of social network hot news public opinion based on knowledge mapping has become the latest research topic.Aiming at the defects of traditional public opinion system,this paper uses the method of text vectorization to improve the accuracy of finding valuable information from the perspective of knowledge mapping.This method has good accuracy and small error in judging the evolution degree of public opinion,the spread of public opinion and the popularity of news.
作者
王丹凤
刘威
WANG Dan-feng;LIU Wei(School of Management,Changchun University,Changchun 130000,China)
出处
《吉林师范大学学报（自然科学版）》
2024年第3期126-131,共6页
Journal of Jilin Normal University:Natural Science Edition
基金
吉林省科技发展计划项目(20240701165FG)。
关键词
舆情发现
知识图谱
话题发现
文本向量化
public opinion discovery
knowledge mapping
topic discovery
text vectorization
