arXiv subjects

Fernando D. O. Castro

Publications and source records attributed to Fernando D. O. Castro.

2 recordsLinked to original sources

arXiv2006

Quantifying social vs. antisocial behavior in email networks

Email graphs have been used to illustrate general properties of social networks of communication and collaboration. However, increasingly, the majority of email traffic reflects opportunistic, rather than symbiotic social relations. Here we use e-mail data drawn from a large university to construct directed graphs of email exchange that quantify the differences between social and antisocial behaviors in networks of communication. We show that while structural characteristics typical of other social networks are shared to a large extent by the legitimate component they are not characteristic of antisocial traffic. Interestingly, opportunistic patterns of behavior do create nontrivial graphs with certain general characteristics that we identify. To complement the graph analysis, which suffers from incomplete knowledge of users external to the domain, we study temporal patterns of communication to show that the dynamical properties of email traffic can, in principle, distinguish different types of social relations.

physics.soc-ph↗

arXiv2005

Improving Spam Detection Based on Structural Similarity

We propose a new detection algorithm that uses structural relationships between senders and recipients of email as the basis for the identification of spam messages. Users and receivers are represented as vectors in their reciprocal spaces. A measure of similarity between vectors is constructed and used to group users into clusters. Knowledge of their classification as past senders/receivers of spam or legitimate mail, comming from an auxiliary detection algorithm, is then used to label these clusters probabilistically. This knowledge comes from an auxiliary algorithm. The measure of similarity between the sender and receiver sets of a new message to the center vector of clusters is then used to asses the possibility of that message being legitimate or spam. We show that the proposed algorithm is able to correct part of the false positives (legitimate messages classified as spam) using a testbed of one week smtp log.

cs.CR↗