Hetero-Labeled LDA: A Partially Supervised Topic Model with Heterogeneous Labels.
Dongyeop KangYoungja ParkSuresh ChariPublished in: ECML/PKDD (1) (2014)
Keyphrases
- unlabeled documents
- topic models
- partially supervised
- text classification
- document classification
- latent dirichlet allocation
- text mining
- text documents
- topic modeling
- text classifiers
- semi supervised learning
- text categorization
- training documents
- unlabeled data
- labeled data
- topic discovery
- latent topics
- generative model
- feature selection
- text data
- latent variables
- training data
- machine learning
- probabilistic topic models
- variational inference
- lda model
- bag of words
- co occurrence
- information retrieval
- naive bayes
- probabilistic model
- multi label
- multi class
- data mining
- knn
- semi supervised
- training examples
- class labels
- multi label classification
- supervised learning
- natural language processing
- information extraction
- data analysis
- learning problems
- vector space