摘要
目前的网络舆情分析系统一般采用输入的关键词进行检索,无法及时发现未知的热点事件。针对这一问题,设计实现了一个具有实用意义的舆情信息分析系统,并将改进的K-MEANS算法应用于文本聚类来自动发现当前的热点主题。运行结果表明,系统可以及时发现热点话题并对事件实时追踪。
For network public opinion analysis system existing general is according to the input keyword retrieval,this method cannot detect hot events unknown.To solve this problem,a analysis system with meaningful public opinion information is designed and achieved,the improved K-MEANS algorithm is applied to text clustering to automatically find hot topics at present.The running result of system shows that it can find hot topics and real-time tracking of events.
出处
《中原工学院学报》
CAS
2014年第6期77-79,84,共4页
Journal of Zhongyuan University of Technology
基金
福建省教育厅B类科技研究项目(JB12487S)
关键词
网络爬虫
特征提取
PCA降维
热点发现
web crawler
feature extraction
PCA dimensionality reduction
hot spots found