OVERVIEW

In this project, I aimed to investigate collaboration patterns among CUNY researchers. Data for CUNY staff working within Computer Science and Mathematics departments was obtained by scraping Google Scholar using SerpAPI. The Neo4j graph database was used for visual inspection and analysis, with Authors encoded as nodes and collaborations encoded as edges. Each node stores the interests of the respective author as an embedded vector, along with their institutional affiliations. Each edge stores the weight of the relationship, calculated as the count of co-authored publications, and the list of years in which these were published. Network analysis was used to predict future collaborations and to investigate author' degree of influence and communities. Future work will explore interests within communities, and patterns of change in these over time.