Privacy in sound-based social networks

In this paper we address the issue of privacy in Online Social Networks (OSN), focusing on those that use environmental sound as a contextual cue for users activity. Through the use of a costume-made research tool consisting of an Online Sound-Based Social Network (OSBSN), we undertook scientific experiments aiming to assess how users deal with the use of sound. Results show that contextual sound is regarded as important and useful for OSN but raises important privacy concerns. In order to deal with this constraint, we propose a system based on the automatic classification of sound environment rather than capturing and sharing actual audio.