Analysis of virtual communities supporting OSS projects using social network analysis S.L. Toral a, * , M.R. Martínez-Torres b , F. Barrero a a E.S. Ingenieros, University of Seville, Camino de los Descubrimientos s/n, 41092 Seville, Spain b Escuela Universitaria de Estudios Empresariales, University of Seville, Avda. San Francisco Javier s/n, 41018 Seville, Spain article info Article history: Received 24 March 2008 Received in revised form 19 August 2009 Accepted 21 October 2009 Available online 25 October 2009 Keywords: Virtual communities Knowledge brokers Social network analysis Open Source Software abstract This paper analyses the behaviour of virtual communities for Open Source Software (OSS) projects. The development of OSS projects relies on virtual communities, which are built on relationships among mem- bers, being their final objective sharing knowledge and improving the underlying project. This study addresses the interactive collaboration in these kinds of communities applying social network analysis (SNA). In particular, SNA techniques will be used to identify those members playing a middle-man role among other community members. Results will illustrate the importance of this role to achieve success- ful virtual communities. Ó 2009 Elsevier B.V. All rights reserved. 1. Introduction VIRTUAL communities can be defined as a social relationship aggregation, facilitated by Internet-based technology, in which users communicate and build personal relationships [37]. They al- low the creation of weak links among geographically dispersed individual who regularly participates in the community. Examples of virtual communities can be found on fields like education [3,41], software development [2] or consumer behaviour [38]. The theoretical background behind virtual communities has been treated by numerous authors. For instance, some authors [34,1] highlight the connection of virtual communities with the so- cial learning theory and communities of practice developed by Wenger [46], while others are focused on their relation with knowledge sharing [24], knowledge creation [26] and innovation models [45]. Another important block of works are devoted to motivation of people participating in online communities [6], the analysis social interactions [39] and participation mechanisms [38,42,43], and practical applications to successful online commu- nities [44]. The social dimension of virtual communities can be justified by thousands of people who freely decide to take part in developing new structures and content [23]. This trend is also confirmed by those organisations which are moving away from traditional struc- tures based on hierarchies, discrete groups and teams towards no- vel structures based on more fluid and emergent organisational forms such as these networks and communities [9]. When people join together in a common enterprise sharing a common expertise, they led to a novel organization form known as community of prac- tice (CoP). CoPs are organized around circumscribed sets of activi- ties and their members develop their own routines, formal and informal ‘‘rules’’, and practices evolve as a result of learning [46]. They do not need to be in direct contact, as nowadays people are in contact using information and communication technologies. Brown and Duguid [7] denominate this kind of communities ‘‘net- works of practice”. This is the case of Open Source Software (OSS) projects, developed under the scheme of a virtual community of support. The user community of an OSS consists of people that, en- abled by Internet, use the software and participate in its develop- ment in some way. The result is geographically distributed individuals who voluntarily contribute to a project by means of the Bazaar model [36,39]. The community of an OSS project can be considered a resource: it is the community that does most of the testing and provides quality feedback, instead of investing huge financial resources to put the software through extensive testing and quality assurance, like a proprietary vendor would do. The major productive assets of a OSS project are the developers, who interact with other developers by forming a small but well or- ganized structure to intensify the learning interactions It has been demonstrated that much of the OSS development is realized by a small percentage of individuals despite the fact that there are tens of thousands of available developers. Such concentration is called ‘‘participation inequality” [24], and it can be explained by the dif- ferent user profiles of open source communities. Moreover, OSS 0950-5849/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.infsof.2009.10.007 * Corresponding author. Tel.: +34 954 48 12 93; fax: +34 954 48 73 73. E-mail addresses: toral@esi.us.es (S.L. Toral), rmtorres@us.es (M.R. Martínez- Torres), fbarrero@esi.us.es (F. Barrero). Information and Software Technology 52 (2010) 296–303 Contents lists available at ScienceDirect Information and Software Technology journal homepage: www.elsevier.com/locate/infsof