SCL seminar by Marija Mitrovic

On 14 June, 2010, Marija Mitrovic from the Department of Theoretical Physics, Institute Jožef Stefan, held a seminar talk entitled "User Communities and emotions on popular stories". Given the social media revolution the field of complex networks, we now have the opportunity to exploit vast amounts of data that can be downloaded from web portals, such as the BBC blog, B92 blog, social networks, Facebook, Myspace, and consumer and product reviews portals like Amazon, IMDb and analyze the behavior of users (whose identity is protected), specifically the emotional content by using a specific set of techniques for text processing and filtering. The emotions are classed as positive, negative or neutral. Social Media data can be represented by bipartite networks and their Projections and systematical detection of communities are based on Eigenvalue Spectral Analysis Method. 

  • DSCF2660
  • DSCF2664
  • DSCF2673