基于K-means-Laplacian的技术演化分析方法研究

Exploration on Technology Evolution Research Method Based on K-means-Laplacian

技术演化研究对企业、政府的科研战略管理具有非常重要的意义。提出了利用K-means-Laplacian算法将专利文本信息与专利引文信息结合起来进行技术演化分析的方法,即首先利用K-means-Laplacian算法将专利文本信息与专利引文信息结合起来对专利文献聚类,然后在聚类结果基础上构建关键词语义网络,最后通过向网络中添加专利申请时间信息,制作技术演化图。旨在将专利两种不同类型的信息结合起来以更加准确的揭示技术演化规律,弥补利用单一信息进行技术演化分析的不足,丰富现有的技术演化分析方法。

Technology evolution research has very important strategic significance for enterprise and government＇s scientific research man- agement. This research proposes a new technology evolution research method, namely using K-means-Laplacian algorithm to combine patent text and patent citation to cluster patents documents, then constructing keywords semantic network based on clustering result, and fi- nally making technology evolution graph by adding patent application time into the network. The paper aims at revealing technology evolution trend more precisely by using two different types of patent information, covering the shortage of previous technology evolution research methods of using single patent information, and enriching existing technology evolution research methods.