Patent Analysis Using Bayesian Network Models
Sunghae Jun, Seung-Joo Lee · International Journal of Software Engineering and Its Applications · 2013
Technology management is important to both companies and nations. Recently, economic competitiveness for both depends on technological prowess. Patents are a representative index of technological competitiveness because a patent document has diverse and detailed information about developed technologies. Hence, patent analysis is an efficient tool for technology management activities such as R&D planning and technology marketing. In this paper, we propose a patent analysis method using Bayesian network models. A Bayesian network is a graphical model representing the relationships between variables. We use International Patent Classification codes as the variables of a Bayesian network model to obtain the technological associations between codes. We verify the performance of our research using retrieved patent documents related to the BMW motor company.