“Uzzi and a team analyzed eighteen million papers from a variety of scientific domains to see whether atypical knowledge combinations mattered. If a particular paper cited other areas of research that rarely, if ever, appeared together, then it was classified as having used an atypical combination of knowledge.”
