A recent study published in Scientometrics presents a two-stage method for identifying breakthrough innovations by analyzing how knowledge is connected and reused. The authors introduce the IPC-PKS algorithm, which measures how closely patents are related by tracking both direct citations and more subtle knowledge links. The approach uncovered 723 breakthrough cases—mostly involving creative combinations of existing knowledge. Follow-up analysis showed that these breakthroughs were strongly linked to real-world impact, based on how often they were cited later. https://link.springer.com/article/10.1007/s11192-025-05276-4
PS - I reached out to the authors, surprised that they hadn’t cited the study published in Nature Communications, which demonstrates that novelty often arises from combining familiar elements. That work introduced the concept of higher-order novelties—first-time pairings of existing components—and employed higher-order Heaps’ exponents to quantify how rapidly such combinations emerge over time. https://19-pacheco-torgal-19.blogspot.com/2025/02/the-hidden-equations-behind-scientific.html
