Prior-aware Composition Inference for Spectral Topic Models.
Moontae LeeDavid BindelDavid MimnoPublished in: AISTATS (2020)
Keyphrases
- topic models
- gibbs sampling
- variational inference
- collapsed gibbs sampling
- dirichlet process
- latent dirichlet allocation
- pitman yor process
- latent topics
- topic modeling
- variational bayesian inference
- bayesian model
- text documents
- text mining
- generative model
- probabilistic model
- latent variables
- prior knowledge
- latent topic models
- bayesian networks
- bayesian inference
- co occurrence
- text corpora
- generative process
- probabilistic inference
- prior information
- em algorithm
- exact inference
- probabilistic topic models
- relevance model
- artificial intelligence