基于专利文献和知识图谱的技术预测方法研究

摘要

This paper proposes an improved analysis method on the basis of the existing methods of patent analysis for tech-nology forecasting.Compared with the existing methods,this novel method has the following innovations:1.using google graph and domain knowledge to create a domain knowledge graph;2.using domain knowledge graph to label the patent documents;3.using social network community evolution research results to predict technology forecasting.Finally,the au-thors select the field of lung cancer for method validation and prediction.The result shows this new method can support the technology forecasting.%在分析现有基于专利文献进行技术预测方法不足的基础上，提出一种基于专利文献和知识图谱的技术预测方法。（1）使用 Google 知识图谱和领域知识创建领域知识图谱；（2）依据创建的领域知识图谱对专利文献赋予标签；（3）引入社会网络社区进化研究成果，基于专利文献标签之间的网络图进行新兴技术预测。以肺癌领域技术预测为例，绘制肺癌领域知识图谱，进行方法验证并预测。验证结果显示，该方法可较好地进行技术预测。