The computation of trust in social networks has attracted the interest of a large number of researchers. Such an interest is explained by the relevant benefits that the usage of trust can bring in multiple application domains, e.g. trust metrics have been incorporated in Recommender Systems to raise their accuracy. Existing approaches to computing trust in social networks can be classified into two categories:
- explicit, in which a user can declare to trust/distrust other users
- implicit, in which user activities are analyzed to infer trust values
In this project we illustrate both the features of explicit approaches and implicit ones and we show how trust metrics can be incorporated in a Recommender System.