论文部分内容阅读
对传统的共词聚类方法进行完善:依据高频低频词界分公式选取高频词;计算粘合力确定每个类别的中心词;对比分析两个时间段,发现主题演变。以医学信息学为例,从PubMed数据库分别下载1999年-2003年和2004年-2008年该学科相关文献,提取主要主题词,进行共词聚类分析,探索医学信息学学科结构的演变过程。
The traditional co-word clustering method is improved: the high frequency words are selected according to the high frequency and low frequency word boundary formula; the cohesion is calculated to determine the center words of each category; and the two time periods are contrastively analyzed to find the theme evolution. Taking medical informatics as an example, the related articles of this discipline from 1999 to 2003 and from 2004 to 2008 were downloaded from the PubMed database. The main keywords were extracted and the cluster analysis was conducted to explore the evolution of the disciplinary structure of medical informatics.