Home // International Journal On Advances in Security, volume 16, numbers 1 and 2, 2023 // View article
Authors:
Igor Jakovljevic
Martin Pobaschnig
Christian Gütl
Andreas Wagner
Keywords: Data Privacy; Open Data; Large Organizations; Clustering
Abstract:
This paper is an extension of our previous work on privacy-protected user clusters identification in large organizations. Oversharing exposes risks, such as improved targeted advertising and leakage of sensitive information. Requiring only the bare minimum of data reduces these risk factors, while simultaneously increasing the privacy of each user. Using anonymized data to find communities opens up new possibilities for large organizations under strong data protection regulations. Although related work often focuses on privacy-preserving community detection algorithms, including differential privacy, in this paper the focus was on the anonymized data itself. Channel membership information was used to build a weighted social graph and groups of interest were identified using popular community detection algorithms. Graphs based on channel membership data resembled interest groups within the network satisfactorily but failed to capture the organizational structure. Furthermore, a statistical evaluation and a user study were conducted to measure the performance of the recommender prototype. The statistical evaluation showed promising results, while the user study yielded mediocre satisfaction of the participants and revealed various potential shortcomings and limitations of the recommender system and user dataset retrieved from the notification system.
Pages: 33 to 43
Copyright: Copyright (c) to authors, 2023. Used with permission.
Publication date: June 30, 2023
Published in: journal
ISSN: 1942-2636