Tracing the origin and diffusion of scientific ideas through automated text analysis

Applying textual analysis tools to the content of the scientific and patent literature can get us pathways to answering fundamental questions on how science and knowledge spread and diffuse. The research therefore strives to apply computer science techniques (machine learning classification models, network theory and natural language processing) to innovation economics, to answer three research questions: (1) How and where does scientific novelty arise? (2) How do novel scientific ideas evolve, compete, and spread throughout science over time? (3) How do novel scientific ideas translate into new technologies?