No CrossRef data available.
Published online by Cambridge University Press: 19 June 2023
Technological life cycles are driven by the changes in the shape and level of innovation, yet innovation rate is not directly observable and is difficult to trace. Innovation measured by patent-citation networks (PCN) cannot be estimated solely by quantity, whereas quality-adjusted quantity measurements are still prone to bias. Our paper complement PCN data analysis with an agent-based simulation (ABS) on networks to uncover the latent innovation that automatically accounts for the quality and quantity components instead of decoupling them. We build dynamic PCN for radio frequency 'CMOS' technology and subsequently develop ABS to replicate underlying innovation network formation. Comparing the real and synthetic data, we isolate latent innovation rate in PCN and by mapping pivotal patent assignees — innovators, we calculate the diversity of innovators in the technology market. Identifying innovation patterns, we show that, early on, innovation structures are less diverse and exploratory, but this grows and matures eventually until value creation become an expensive endeavor. Contrary to what is observed, we show how the abundance that appears are less significant publications mostly driven by exploitation in the technology market.