Improving Topic Models with Latent Feature Word Representations

Dat Quoc Nguyen, Richard Billingsley, Lan Du +1

Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature vector representations of words have been used to obtain high per…