PeoplePivot – Temporal and Spacial navigation of co-authorship data

The DevCSI developer challenge at #repofringe11 was focused on the events theme “repositories building bridges”. As people of repositories Matt, Andy and myself thought that what we wanted to do was make a tool for exploring co-authorship over the course of a persons career.

We came up with a simple visualization which enabled the user to see how closely people were working together. The closeness of two people on the graph represented the closeness represented how closely they had worked together. We defined this closeness of a function of how many papers they had co-authored with and how many people they had co-authored with. Our rationale was that if a person is co-authoring a lot of papers with a small circle of authors those authors are working closely together and this is represented by their closeness on the graph.To capture a researchers career we generate a graph for each year they have published and then transition from year to year moving, adding and removing the nodes to reflect their career over time.

A user can navigate the graph by clicking nodes to see the co-author relationship. The user can also pivot through the author space by selecting a node and making it the new focus of the graph. This ability to re-pivot the data around a different people allows users to see rich interconnections in an interactive discovery of the co-authorspace.

The original version of PeoplePivot is no longer available – but we’ll blog about it again in the future.