Dr. Anna Latour @anna@mathstodon.xyz
Public

My student did a Science!

One week from now, Denisa Arsene will present the her BSc thesis research at the 14th International Conference on Complex Networks and their Applications, held in Binghamton, New York, USA!

More details and preprint: latower.github.io/posts/2025/1

I'm super proud of her, and I'm super happy that I got to do this project with her and two of my former colleagues from Leiden University: Rachel de Jong and Frank Takes!

Poster presenting the highlights of the research in the corresponding paper. Eyecatcher is a big vertical orange banner with the text: 'How to share research-relevant social netowrk data, while preserving privacy? SANA Simulated Annealing for Network Anonymization. Minimal changes. Maximal privacy. Up to 18 times more anonymized nodes. Same speed. Same data utility. It features diagrams that illustrate a type of attack model that is based on characteristics of the direct neighbourhood of nodes. It also has an example plot of how a simulated annealing approach converges to a particular level of anonymity. It then has a bunch of plots that give more details on running time, level of anonymization, and how SANA and the other algorithms perform in terms of data utility. The takeaways are that SANA provides fast and better anonymization and that it scales better with input size than the other algorithms studied here. Next challenges are to tackle multi-objective anonymity and utiliyt, and dynamic (real-time) anonymization.
0
0
0

If you have a fediverse account, you can quote this note from your own instance. Search https://mathstodon.xyz/users/anna/statuses/115650174946086904 on your instance and quote it. (Note that quoting is not supported in Mastodon.)