General Information
NameProf. Rami Puzis
DepartmentDepartment of Software and Information Systems Engineering
Emailpuzis@bgu.ac.il
Personal Web SitePersonal Web Site
Academic RankAssociate Professor
FunctionM.Sc committee - chairman, Department of Software and Information Systems Engineering


Bibliography

  Scholar DBLP
   

Research Interests

 
  • My main research interests lie in complex networks analysis, an area of study that involves a variety of domains ranging from the analysis and investigation of social networks to computer security. In the field of computer security, my focus has been on the analysis of computer communication networks to protect end-users and critical infrastructure facilities from malicious attacks and incidents via cost effective deployment of monitoring and filtering devices. In the field of social network analysis I am interested in link prediction algorithms, models representing the spread of media content, opinion or epidemics, identification of key players etc.
  • Pinpointing the most significant nodes in social or computer networks requires developing efficient algorithms in two domains: search and centrality computation. First domain includes optimization algorithms for finding the group of key players.
  • One of the most challenging tasks in the second domain is speeding up the computation of centrality indices. An algorithm for fast successive computation of Group Betweenness Centrality (GBC) can be found here. After a pre-computation step, the time required to compute GBC of every given group does not depend on the network's size. Thus, the algorithm can be used for finding the most central groups even in large-scale networks.
  • Centrality computation speedup includes various techniques ranging from transforming the network structure, through centrality approximation, to parallel algorithms executed on general purpose graphical processing unit (GPGPU).
  • Besides increasing the efficiency of centrality computation algorithms, it is very important to device centrality measures that are suitable for a variety of real world applications. Routing Betweenness Cewntrality is a generalization of betweenness centrality that can be used with an arbitrary (loop free) routing strategy.
   

Research Projects

 
  • Social Network Crawling for Targeted ‎Group ‎and Organization ‎Data ‎Collection.
  • Reconstruction of Hidden or non-existent Identities Through Automatic Social Network Crawling.
  • Optimal Network Monitoring for detecting APT
  • Evaluation Environment for Simulating Cyber Attacks.
  • Detecting Hidden Links and Nodes in Public Social Networks