Scalable Inference for Latent Dirichlet Allocation

James Petterson, Tiberio Caetano

We investigate the problem of learning a topic model - the well-known Latent Dirichlet Allocation - in a distributed manner, using a cluster of C processors and dividing the corpus…