Exploiting the value of class labels in topic models for semi-supervised document classification.
Hossein SoleimaniDavid J. MillerPublished in: IJCNN (2016)
Keyphrases
- document classification
- class labels
- topic models
- semi supervised
- labeled data
- text documents
- classification algorithm
- unlabeled data
- text mining
- supervised learning
- text classification
- topic modeling
- generative model
- text categorization
- semi supervised learning
- latent dirichlet allocation
- active learning
- latent variables
- co training
- probabilistic model
- multi label
- pairwise
- training data
- transfer learning
- training set
- co occurrence
- unsupervised learning
- feature set
- training examples
- text classifiers
- document clustering
- gibbs sampling
- markov networks
- web documents
- naive bayes
- feature selection
- nearest neighbor
- knn
- data points
- feature vectors
- similarity measure
- information retrieval
- data mining