共词矩阵在聚类结果分析中的作用

Role of co-word matrix in analysis of clustered words

共词聚类分析法的聚类过程存在排斥性和敏感性缺点,使分析的科学性受到质疑。共词矩阵是词对共现频率的总集,对共词聚类结果(类团)有决定性的影响,在类团的分析中能弥补聚类算法的不足。通过实例研究,认为共词矩阵有助于对类团成员的修正及对类团成熟度的判断。

The scientificalness of co-word clustering analysis is questioned due to its exclusiveness and sensitivity during the clustering process. Co-word matrix, which is a collection of co-word frequencies, has a decisive effect on the clustered words and can counteract the disadvantages of clustering algorithm. Based on the case study, we hold that co-word matrix contributes to the correction and maturity judgment of clustered words.