VISUALIZING COLLABORATIONS Coauthorship Networks
The Need Potential Merger of 2 Colleges: CFANS: College of Food, Agriculture & Natural Sciences CBS: College of Biological Sciences
Informatics/Data Services Specialist Steven Braun Created visualizations to show the co-author relationships between the two colleges
What Data is Represented? Data harvested from Publications indexed in Scopus database Publication co-authorship relationships Research concept/keyword profiles for researchers Biographical information via UMN’s PeopleSoft (department affiliation(s), title, etc.)
What Can We Analyze? Collaboration = Coauthors on publications Collaboration networks illustrate connections within/across colleges, disciplines, and core domains of research interests Networks show who: –work together frequently –are high-productivity hubs –have unrealized opportunities.
Hierarchical Network View
Bundle
Hierarchical Concept Network
Cautions & Limitations Extrapolating collaboration networks backwards in time with old publication data using knowledge of current college affiliation may produce false positives Publications are harvested from Scopus only; database is extensive but not comprehensive Current configuration of PeopleSoft and limitations in how much data on researcher title/ranks and affiliations can be ingested by may mean some missed connections
REACHING OUT Featuring Centers and Institutes
Libraries staff manually added Research Interest Keywords
Really just a search result
CAREI Visualization Example Kyla Wahlstrom, PhD Senior Research Fellow
Finding the Balance The Libraries can –Develop the data model needed –Train Center staff But then the Center must –Work with the Libraries to determine their needs –Commit staff time to maintaining their presence