摘要
基于舆情事件的词汇关联分析,既是面向网络舆情的情报研究中的一项关键技术,也是保证和提高网络舆情分析质量的一个重要途径。文章研究基于词跨度的关键词获取算法,对候选关键词进行权重计算。研究计算词汇之间的共现率算法,通过限定范围和结果组配的方法识别词汇间的关系。实验测试取得了良好效果,对于提高网络舆情事件分析的质量有重要意义和应用价值。
The words co-occurrence analysis on public opinions is not only a key technology in network public opinion analysis, but also an important way to ensure and improve quality of network public opinion analysis. This paper studies the acquisition algorithm based on word span keywords which filters out unrelated words. Word frequency is computed and the location of candidate key- words is marked. The computing vocabulary between the co-occurrence rate algorithm is studied through limited scope and results of group with methods to identify the relationship between words, and a set of keywords are drawn. Experimental test has achieved good results and verifies the significant application value for improving the quality of public opinion analysis.
出处
《信息工程大学学报》
2014年第1期105-110,共6页
Journal of Information Engineering University
基金
科研基金资助项目
关键词
舆情事件
关键词抽取
关联分析
public opinion event
keyword extraction
correlation analysis