Tracking innovation diffusion: AI analysis of large-scale patent data towards an agenda for further research

Recent advances in AI algorithms and computational power have led to opportunities for new methods and tools. Particularly when it comes to detecting the current status of inter-industry technologies, the new tools can be of great assistance. This is important because the research focus has been on how firms generate value through managing their business models. However, further attention needs to be given to the external technological opportunities that also contribute to value creation in firms. We applied unsupervised machine learning techniques, particularly DBSCAN, in an attempt to generate a macro-level technological map. Our results show that AI and machine learning tools can indeed be used for these purposes, and DBSCAN is a potential algorithm. Further research is needed to improve the maps and to use the generated data to study related phenomena including entrepreneurship.