Extracting Discriminative Concepts for Domain Adaptation in Text Mining

One common predictive modeling challenge occurs in text mining problems is that the training data and the operational (testing) data are drawn from different underlying distributions. This poses a gre