The Internet has brought in the last two decades an explosion of online communities, in some cases spanning over a billion participants - communities that are qualitatively different from traditional ones both because of intrinsic properties and because of their massive scale. Their analysis, and the analysis of the analysis tools themselves, remains a fertile ground for research (and one with a lot of money in it!) at the crossroads of many disciplines.

Recommender systems (e.g., Amazon, LinkedIn, ...) and reputation systems (e.g., Tripadvisor, Booking.com, ...) have the purpose of providing advices for decision making in e-commerce and online service environments. They are similar in the sense that both collect data of members in a community in order to provide advice to those members. However, there are also fundamental differences. The purpose of recommender systems is mainly to generate suggestions about resources that a user a priori is not aware of but is possibly interest in.

Research activities:

Computational models for expression communication can contribute to the definition of innovative human-computer interfaces, based on devices capable of understanding and responding to the user’s affective state and behavior.

Research activities:

Research activities:

• Development of innovative techniques for spatial audio synthesis with particular attention to binaural audio synthesis based on Head-Related Transfer Functions (HRTFs) and real-time audio rendering.

• Development of the "Mixed Structural Modeling (MSM)" framework for the design of HRTFs, combining synthetic modeling paradigms with HRTF selection techniques.

• Customization of HRTFs, based on individual anthropometric data, which are used to either fit model parameters or to select recorded components within a dataset of available responses.