D-LDA: A Topic Modeling Approach without Constraint Generation for Semi-defined Classification.
Fuzhen ZhuangPing LuoZhiyong ShenQing HeYuhong XiongZhongzhi ShiPublished in: ICDM (2010)
Keyphrases
- topic modeling
- topic models
- latent dirichlet allocation
- text classification
- text mining
- modeling framework
- feature extraction
- topic extraction
- document classification
- pattern recognition
- latent topics
- latent semantic analysis
- machine learning
- text corpora
- feature selection
- collaborative filtering
- variational inference
- probabilistic latent semantic analysis
- support vector
- classification algorithm
- text documents
- dimension reduction
- generative model
- face recognition
- hierarchical bayesian model
- gibbs sampling
- probabilistic topic models
- latent variables
- search engine
- training set
- support vector machine svm
- unlabeled data
- feature space
- natural language processing
- probabilistic model