Export Reviews, Discussions, Author Feedback and Meta-Reviews

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper proposes to learn a text model APM (Inouye+, 2014) for large datasets by alternating minimization. APM is an admixture of Poisson random fields on words, thus like an LDA where topic distributions are replaced by Poisson random fields. As such, learning possible interactions between words is hard for large vocabularies. Authors propose an EM-like algorithm where Poisson random field parameters are optimized in the M step.