Gaussian Copula Marginal Regression Modeling for Technology Analysis

Abstract

Abstract-Understanding technological relations between patent technology keywords is an important task for building research and development (R&D) policy of nation and company. Many researches have been actively conducted on this research subject, and various approaches to technology analysis were studied in the field of technology management. Most of the methods of technology analysis were based on patent documents related to target technology, because patent contains diverse information on developed technologies. So the patent keywords extracted from patent documents are valuable sources for technology analysis. The structured patent data become a matrix consisting of patent (row) and keyword (column), and each element of the matrix is frequency value of the keyword occurred in each patent. In this paper, we propose a method of technology analysis using Gaussian copula marginal regression (GCMR) model, and use the R data language for patent analysis by the GCMR. In addition, we carry out a case study to show how this study could be applied to real problem. This research contributes to various R&D planning of nation and company.