Unsupervised Transfer Classification: Application to Text Categorization

We study the problem of building the classification model for a target class in the absence of any labeled training example for that class. To address this difficult learning problem, we extend the id