Analyzing social networks

    Virtual social networks are used by hundreds of millions of people, and are rapidly growing. In parallel the need for analyzing these networks also keeps growing. We try to solve two separate problems within this research field.

    We are modeling the flow of information within a network, based on Last.fm user network and the users' habits of listening to music. We are investigating how users are affected by their friends' behavior, and their preference of music artists. We are looking for a mathematical model to describe these effects, and will try to use our findings to give more precise recommendations to users.

    In another project we try to identify the same users in different networks. This is a special case of entity resolution, where the identification of entities is based not only on their data attributes, but also on their environment within a network.

    Languages