Prediction of Technology Convergence on the Basis of Supernetwork

Given the significance of technology convergence for innovation and launching new products, accurately predicting technology convergence is critical to the pursuit of technological innovation. Although previous studies have been proposed to predict technology convergence, the framework using single-layer international patent classification (IPC) networks neglects the potential connections between components. This issue leads to the prediction that IPC nodes in separate components will not overlap. This study utilizes IPC, patentee, and topic information to construct a technology supernetwork model consisting of “Topic-IPC-Patentee” three layers, aiming to overcome the issue of IPCs located in different components being unable to connect. Based on the supernetwork, we adopt a link prediction method based on superedge similarity to predict the technology convergence and a case analysis on the field of gene editing is conducted. According to experimental results, we observe that the technology supernetwork model can effectively predict the convergence of gene editing technologies, demonstrating robust predictive performance. The main contribution of this research is to provide a methodology that accurately predicts technology convergence, which can help firms deploy their research and development (R&D) strategies, policymakers develop insightful policies, and investors capitalize high-value projects.