Research on MDS-based visual clustering in patent texts
基于多维标度法的专利文本可视化聚类研究*

In order to acquire the information need from the current large amounts of patent texts and make them visualized to people quickly and exactly, this paper proposed a kind of visualized clustering method in patent texts based on MDS through researching the correlative theory and key technology of text mining, combined with the character of patent texts and existing analysis methods. It also made some relative analysis and discussion about the application of this method. The experiments show that this method displays the current distribution of patent texts application topics well; it can also effectively support the decision in management and technology innovation.