A genetic search of patterns of behaviour in OSS communities

This paper proposes the identification of patterns of behaviour of open source software (OSS) communities using factor analysis and their social network analysis (SNA) features. OSS communities can be modelled as a social network in which nodes represent the community members and arcs represent the social interactions among them, and factor analysis is able to provide the factors that explain the latent patterns of behaviour. Due to the complexity of the problem and the high number of SNA features that can be extracted, this paper proposes a genetic search of an optimum subset of indicators leading to a group of latent patterns of behaviour maximizing the explained data variance and the interpretation of factors. Obtained results illustrate the feasibility of the proposed framework to extract relevant information from a large set of data.

► Representation of OSS communities as social networks.
► Genetic search of the optimum subset of indicators representing patterns of behaviour in OSS communities.
► Classification of OSS communities into four patterns according to Social Network Analysis features.
► Representation of OSS communities in a bidimensional space given by size and hierarchy.