Joint probability models are useful for unsupervised learning ("knowledge discovery") as well as supervised learning tasks such as where we want to predict multiple correlated outputs ("collective classification"). One approach to representing such models is to explicitly model the correlations using a graphical model, where we add edges between correlated variables (nodes).   Another approach is to implicitly model the correlations via a set of latent common factors; this induces dependence between the visible variables without the need to add explicit edges between them.