31b3b31a1c2f8a370206f111127c0dbd-Reviews.html

"NIPS 2013 Neural Information Processing Systems December 5 - 10, Lake Tahoe, Nevada, USA",,, "Paper ID:","1009" "Title:","When Are Overcomplete Topic Models Identifiable? First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper gives a sufficient condition for a unique recovery of topic-word matrix of n-persistent overcomplete topic models by a tensor decomposition of moments. In the overcomplete regime, the number of topics are more than that of words. Thus, it is impossible to uniquely recover the relation between topics and words.