Transferring Topics: A Framework for Lowering the Costs of Supervised Learning in Text Analysis

• In Event: Text as Data: Methods and Applications

Brief Overview

We introduce a framework for lowering the costs of building a training set by leveraging already labeled data from different sources. We apply this approach to the task of assigning the Comparative Agenda Project’s major topic codes.

Authors

• Kelsey Shoub, University of South Carolina

• Alison Craig, University of Texas - Austin