Mobility transition in the automotive industry in Germany and Japan: A patent network analysis (data and graphs)

Patent documents offer insights into the countries dedicating resources to specific technologies, enabling the identification of potential innovation trends, particularly emerging technologies, and the development of innovation capabilities. International Patent Classification (IPC) codes are assigned to patents, much like keywords are assigned to scientific articles. For this study, two IPC code co-occurrence networks are constructed. Both IPC co-occurrence networks are undirected and consist of nodes (IPC codes) and weighted edges (the number of times IPC codes co-occur). IPC codes are co-occurring if they are assigned to the same patent. Codes are included in the analysis if they occur at least two times and are part of the largest connected component of the network.

The networks are based on the patent data of the leading Japanese and German car manufacturers. The main search query (223,688 patents) is narrowed down to patents with application dates from 2012 to 2023 (183,706). The Japanese dataset (9,685 patents) limits the main query to the three largest Japanese car manufacturers: Honda, Nissan, and Toyota (including subsidiaries: Daihatsu and Hino). The German dataset (1,381 patents) limits the main query to the three largest German car manufacturers: BMW, Mercedes-Benz, and Volkswagen (including subsidiaries: Audi and Porsche). The patent data are downloaded from the PATENTSCOPE database of the World Intellectual Property Organization. For network analysis, we use Gephi and VOSviewer. The PATENTSCOPE data are imported into VOSviewer and formatted as co-occurrence networks. Gephi is used to calculate diameter-based centrality and modularity class. The graph layout is ForceAtlas2.