This article introduces some of the most popular techniques of data integration, that allow the combination of information coming from various sources. The illustration focuses, in particular, on the Bayesian generalized Heckman methodology and on the data calibration methodology based on vines and nonparametric Bayesian networks.