Research On Visualization Of Patent Citation Analysis

At present, the world technology competition is becoming more and fiercer, the patent strategy research is carried out by the enterprises in various countries, and the core is the patent information analysis. As an important part of patent analysis, patent citation analysis refers to the analysis of patent citation and citation. Through citation analysis, we can find the core patents, technology trends and potential competitors in the specific field. Due to the huge amount of patent information data, efficient and rapid patent citation analysis must be based on professional and visual analysis tools. At present, the commercialization of patent analysis tools are more, but most are just the primary level of patent information statistics and analysis tools, such as patent citation analysis and visualization tools are not perfect. The purpose of this study is to improve the traditional citation analysis visualization method, in order to improve the multi-level patent information reveal, patent analysis of readability and interactivity.Firstly, the paper makes a deep analysis on visualization methods and research status of visualization technology and patent citation analysis. On this basis, to solve the problems that exists in the traditional hierarchical data visualization method of Hyperbolic tree, including the limited capacity of revealing multiple patent information, dense node region occlusion and limited display capacity of related details of the patent information, an improved Hyperbolic tree patent citation analysis visualization method is proposed. This method is mainly to improve the transfer from the three aspects of hyperbolic tree layout, interactive and dynamic, the traditional Hyperbolic tree and fisheye technology combined, and adding seven kinds of dynamic interactive technology. The improved visual model can not only reveal the overall layout of the citation information, but also show the details of the local node.Secondly, according to the fact that the patent citation network has large amount of multi-dimensional data, we use a simplified icon to turn the regular, representative points and edges into meaningful icons, and the corresponding graph detection method is designed for the simplified icon. The method can reduce the scale of the citation network data, and can improve the readability of the citation network. At the same time, it can also reveal the evolution of patent technology in different years, different invention main body, and then it will be conducive to the discovery of the technological evolution path of massive patent.Finally, the design and implementation of a visualization tool for patent citation analysis, this tool is for the United States patent, including search and acquisition module of citation information, data cleaning module, analysis module of citation information visualization.In the tool, Python is used for patent data acquisition, and through regular expressions, patent information is extracted and loaded to a SQL Server database, and simple statistical charts, the Hyperbolic tree, simplified citation network visualization method are provided for the user,which can greatly reduce the workload analysis of the user and improve the effect of visual expression.