Supporting Network Formation through Mining under Privacy Constraints

Abstract

Single professionals and small companies come together and form virtual communities to compete with global players. In these collaboration networks, the actual business partners are discovered and alliances formed on demand. However, it is impossible for single members to keep track of the dynamics in large-scale networks. With the wide adoption of service-oriented architectures (SOA), interactions between partners have become observable. Monitoring collaborations enables the inference of social relations and the identification of successful partner compositions. Measuring the quality of social relations, such as the degree of trust based on the success of past interactions, are a powerful means to support the formation of alliances. However, by applying monitoring, also privacy concerns arise. In this paper we deal with concepts and tools to support group formations. We consider the trade-off between the benefits of sharing personal profiles and accounting for privacy concerns of the individual network members.