Recherche

Online Social Networks (OSNs) are the breeding ground for a large number of different services. These services may be offered by the OSN itself; this is the case for example of recommendations for Facebook, job recruiting for LinkedIn, collaborative working tools for Atlassian, Spigit or Yammer, stock trading for FX Copy… In other cases, services are offered as third-party applications taking advantage of the OSN platform. Among the most popular examples in the Facebook ecosystem there are Band Profile and ONErpm to promote musicians, Cause for participative democracy support, Where I’ve been to share travel informations, Angry Birds Friends and Bejeweled Blitz (and many others) for gaming.

Our goal is to study how services in OSNs can be efficiently designed and managed. This research requires to answer 3 main questions:

  1. How can the topology of an OSN be discovered? Many services need or can take advantage of some knowledge of the network structure that is usually not globally available and in any case changes continuously due to structural dynamics.
  2. How does services’ adoption spread across the OSN? On the one hand, the popularity of a service is often determined by word-of-mouth through the links of the OSN, on the other hand, the service may contribute to reshape the structure of the OSN (e.g. by creating new connections).
  3. How do different services compete for the finite attention and money of OSN users?

In particular our purpose is to provide analytical models (corroborated by simulations and experiments on real networks) to understand such complex interactions.