A distributed tool for online identification of communities in co-authorship networks at a university

Abstract

Most universities have their public repositories of scientific publications available online. The data is made available raw or by department listing and does not provide the network of co-authorships that implicitly emerges from scientific collaborations among different departments. Sometimes, the network of co-authorships is computed within the institution, via standalone applications that have few or no functionalities to explore the structure of collaborations. The possibility of searching online and managing the network of scientific communities in the institution is a matter of management efficiency, both for the institution itself and other external collaborators. This paper explains a distributed architecture and a tool that uses data from an online institutional repository. The tool calculates and puts available online the co-authorship network that identifies research communities according to different algorithms. The tool reflects and identifies the emergent structure of communities, graphically analyses communities, exports, reports and follows up with the evolution of communities in time.