Exploring ideation: Knowledge development in science through the lens of semantic and social networks

In this paper, we explore changes in both structural and semantic characteristics of a scientific social network. We trace the emergence of knowledge, what we refer to as ideation, through publication data from two conferences in a sub-field of Computer Science. Social network analysis is used to determine structural characteristics of the co-authorship networks, and we perform semantic network analysis on title words of articles to trace content of topics over time. We find that the emergence of new topics is accompanied by a tendency toward less dense and transitive, but more evenly distributed, social networks. We show that ideation is fostered in a relatively loose environment, as the field does not cling to one particular topic with all authors contributing to it. On the contrary, there is enough 'slack' for new knowledge topics to emerge and consolidate over time. Findings contribute to research on social network antecedents, innovation and science studies.