Research On The Dmc-based Patent Tech Mining

National policy highly emphasizes that innovation is the greatest power of leading the development of science and technology,and the innovation of technology has become the strategic support to improve the social productivity and the overall national strength.Research on the field technical status can help grasp competitive intelligence and provide information for business decisions.In the early,research based on bibliometric explored the technology trends through literature statistical analysis.Nowadays,with the development of text mining technology,tech mining based on patents provides a new perspective for technical analysis.This paper exploits the research status both at home and abroad,and analyzes related methods to illuminate the existing problems in current research.The problems include improper feature selection and the uncomprehensive measurement of the correlation between technologies.Therefore,this paper works for searching proper solutions to solve these problems above,and here lists the specific work:Firstly,this paper studies the theory,models and key technologies of text mining and SNA,and also understands the characteristics and functions of patents,and then states the integration of text mining and SNA in patent analysis.After that,this paper illustrates that the Derwent Manual Code(DMC)can be chosen as the feature of patents.Compared to domain ontology,the links between technologies can be measured by evaluating the concept semantic similarity between the codesThen,this paper introduces the concepts and the process of tech mining,and proposes the method of patent tech mining based on DMC.This paper uses DMC to represent technology,and measure the inherent links between technologies by evaluating the concept semantic similarity between the codes,and also measure the technology co-occurrence by evaluating the code co-occurrence in patents.And then introduces both of them into traditional tech mining to avoid the uncomprehensive results under incomplete standard.At the same time,this paper introduces the degree centrality and the betweenness centrality into technical analysis based on maps to measure the importance of technologies.At last,we study on robot to test the feasibility and rationality of the method.