Ni zapisov
ITU News: Forging paths to 5G
Avtorji: Haoulin Zhao, Dongmyun Lee, Francois Rancy, Stephen M. Blust, Sergio Buonomo, Chaesub Lee, Peter Ashwood-Smith, Giovanna Carofiglio, Marc Cohn
Novel Cross-Layer QoE-Aware Radio Resource Allocation Algorithms in Multiuser OFDMA Systems
Avtorji: Rugelj, M., Sedlar, U., Volk, M., Sterle, J., Hajdinjak, M., Kos, A.
WEB on TV: Cross-National User Study for Designing Better User Experience
Avtorji: Emilija Stojmenova, Jože Guna, Daniela Hauswirth, Roland Winkler, Manfred Ninaus and Matevž Pogačnik
     Povečaj pisavo   Pomanjšaj pisavo

User behavior analysis based on Identity management systems’ log data

Avtorji: Borut Rožac, Radovan Sernec, Andrej Košir, Andrej Kos

Identity Management Systems (IdM systems) are repositories where users’ security credentials are kept and managed. One of many tasks performed by IdM system is an authentication and authorization of users while they are using applications. In case where IdM systems are organized as central repository, which means that authentications and authorizations for different applications are carried out on one central IdM system. This case is common for enterprises with more than 50 employees where various aspects of users’ behavior can be analyzed. By applying machine learning method SVM (Support Vector Machines) on IdM system log data with application on searching employees that are acting enthusiastically within working day, we showed that analysis of IdM systems log data can be successfully applied for analyzing employees behavior within an enterprise.

Za ogled vsebine si prenesite PDF datoteko.

Prispevek je objavljen v Zborniku 21. mednarodne Elektrotehniške in računalniške konference ERK 2012.


Ključne besede: identity management, SVM