Identifying potential technological spin-offs using hierarchical information in international patent classification
提出一种利用国际专利分类层次信息构建假设共现网络的方法，用于预测技术衍生事件，案例显示分类模型F1值超0.94，回归模型调整R方超0.73。
As the relationships among technologies become more complex, technological convergence is occurring in many fields, resulting in technological spin-offs in which technologies born in an industry are used in unexpected fields. Predictions of these events, such as patent citation analysis and International Patent Classification's (IPC) association analysis, are used to propose a method. Previous studies have not taken into account the hierarchical structure of technologies. In this study, we propose a hypothetical co-occurrence network in multiple technology layers using IPC's hierarchical information to show that features of different hierarchies can contribute to predictive performance and interpretation. Patent information on carbon fiber reinforced plastics and functionally graded material is extracted from patent database, Thomson Innovation, for a case study. The results show that an F1 measure of classification model exceed 0.94 and an adjusted R2 of regression model exceed 0.73. The existence of key common IPCs, which occurred in a different layer than the spin-off prediction target, allowed us to identify the technology fusion behind each specific example. The identification and prediction of technological spin-offs can contribute to research and development strategies and the development of potential business partners.