Effects of Social Networks On Anonymity Systems

Anonymity, social network, email, mix

One of the biggest problems in the research of the communication anonymity systems is missing real-world data. Large number of results related to anonymity systems assumes that use behaviour can be described with simple probabilistic models. This paper shows some basic interesting characteristics of email communication we have identified in a log of SMTP traffic and demonstrates that user behaviour cannot be easily described with simple models. Basic properties of existing social networks are described. We analyse the traffic from the two main viewpoints: intra-domain and complete communication. We created subsets of communication within university faculties to analyse social networks as we have all the traffic that took place in these domains. The second viewpoint is represented with an analysis of all traffic initiated by users from these email domains.