Discovering Scientific Communities using Conference Network

This paper presents an algorithm and a tool for discovering scientific communities. Several approaches have been proposed to discover community structure applying clustering methods over different networks, such as co-authorship and citation networks. However, most existing approaches do not allow for overlapping of communities, which is instead natural when we consider communities of scientists. The approach presented in this paper combines different clustering algorithms for detecting overlapping scientific communities, based on conference publication data. The Community Engine Tool (CET) implements the algorithm and was evaluated using the DBLP dataset, which contains information on more than 12 thousand conferences. The results show that using our approach it is possible to automatically produce community structure close to human-defined classification of conferences. The approach is part of a larger research effort aimed at studying how scientific communities are born, evolve, remain healthy or become unhealthy (e.g., self-referential), and eventually vanish.