Congfeng Cao (Feng) is a PhD student working on methods to represent scientific and technological innovation. Scientific publications, patents and citations among them are regarded as a tangible source of data to explore innovation. Feng’s project proposes to further our understanding of what characteristics innovations in science and technology possess, and how they disrupt the previous state of the art creating a new equilibrium. Feng uses methods from applied machine learning, especially natural language processing and representation learning on graphs, to exploit such large-scale contents. What is more, this work is informed by theories of scientific progress (e.g., Kuhn’s paradigm shifts, Becher and Trowler’s tribes and territories) to be tested empirically.
Feng’s background is in machine learning and deep learning. He has worked at a number of multinational companies for over 5 years, as a senior algorithm engineer and team leader.