Generating Anonymous Social Networks for Analysis Tool Development

C. Lofdahl

Proceedings of the 81.2 Military Operations Research Society (MORS) Symposium, Alexandria, VA, (June 2013)

Social media use has exploded over the past decade and with it the availability of data that can provide insight into a range of behavioral questions. Social media data however is so voluminous, complex, and hard to interpet that it cannot be reviewed directly but instead must be analyzed with the aid of computer-based tools. Developing these computational tools requires having social media datasets available on which to test their analytic algorithms, but privacy concerns limit the distribution of actual social media data. Originally such data was anonymized by removing sensitive personal data, but increasingly successful deanonymization attacks leveraged other networks with similar information to reveal personal information. Anonymous networks provide a possible solution because they are generated by algorithms rather than from actual social media, so they are not susceptible to deanonymization attacks. This paper discusses an Agent-based System Produced Emergent Network (ASPEN) framework that employs social science theory and computer simulation to produce anonymous networks for use by social media tool developers.

For More Information

To learn more or request a copy of a paper (if available), contact C. Lofdahl.

(Please include your name, address, organization, and the paper reference. Requests without this information will not be honored.)