Additive Bayesian Network: Exploring The Latent Networks From Textual Responses and Scores

• In Event: Network Models

Abstract

This study used the additive Bayesian network (ABN) model to build latent networks between latent attributes learned from a diagnostic classification model and latent topics learned from a latent Dirichlet allocation model. Preliminary results showed that the ABN is useful for detecting the complex relationships between these latent variables.

Authors

• Constanza Mardones-Segovia, UNIVERSITY OF GEORGIA

• Shiyu Wang

• Allan Cohen, UNIVERSITY OF GEORGIA

• Xiaoming Zhai, University of Georgia