V. Herskovic et al. (Eds.): CRIWG 2012, LNCS 7493, pp. 73–88, 2012. © Springer-Verlag Berlin Heidelberg 2012 Designing the Software Support for Partially Virtual Communities Francisco Gutierrez 1 , Nelson Baloian 1 , Sergio F. Ochoa 1 , and Gustavo Zurita 2 1 Computer Science Department, Universidad de Chile Av. Blanco Encalada 2120, 3rd Floor, Santiago, Chile {frgutier,nbaloian,sochoa}@dcc.uchile.cl 2 Control Management and Information Systems Department, Universidad de Chile Diagonal Paraguay 257, Santiago, Chile gzurita@fen.uchile.cl Abstract. Designing software platforms to support the activities of partially virtual communities (PVC) is a challenging task since the supporting services must evolve continually according to the community evolution. Moreover, unsuitable supporting services usually lead the community to its demise. Therefore, these platforms must count on a flexible architecture that provides suitable services as a way to support interactions among community members, and thus contributing to keep the community sustainability. This article proposes a software architecture that helps software designers to address this challenge. Such a model can be used not only to ease the architectural design process, but also to evaluate already implemented PVC supporting systems. The article also shows a preliminary evaluation of both roles of the proposed model and discusses the obtained results. Keywords: Social system architecture, software architecture, partially virtual communities, supporting systems. 1 Introduction Over recent years, social computing has become present in many aspects of our daily activities. Although virtual communities have been present in several scenarios for some time, the recent rise of social computing systems has helped spread and diversify them. Several taxonomies have been proposed to classify these people associations [16, 28]. This article considers just one of these types that we have called Partially Virtual Communities (PVC) [11]. In such communities, members have the opportunity to interact frequently through both a virtual and a physical space. Examples of PVCs are the communities of a university course or people in a small neighborhood. Membership in these communities is quite stable, meaning that few people join or quit these communities. PVCs depend on a certain personal interaction and knowledge among their members. Therefore, when two people decide to interact through the virtual space (i.e. the community supporting system), they know each other, and such contextual information (i.e. the mutual knowledge) allows them to appropriately interpret others’ contributions. When a member makes a commitment, the rest of the participants know (or estimate) how trustworthy that commitment is,