Text classification with relatively small positive documents and unlabeled data.
Fumiyo FukumotoTakeshi YamamotoSuguru MatsuyoshiYoshimi SuzukiPublished in: CIKM (2012)
Keyphrases
- text classification
- unlabeled data
- labeled data
- text documents
- unlabeled documents
- document classification
- training documents
- text classifiers
- text data
- labeled data for training
- co training
- text categorization
- semi supervised learning
- labeled training data
- labeled and unlabeled data
- text mining
- feature selection
- semi supervised
- bag of words
- labeled examples
- naive bayes
- class labels
- machine learning
- document collections
- multi label
- semi supervised classification
- training data
- transfer learning
- positive and negative
- document clustering
- training set
- information retrieval
- active learning
- knn
- unsupervised learning
- positive and unlabeled examples
- supervised learning
- sentiment analysis
- web documents
- training examples
- neural network
- small number
- learning algorithm
- data mining
- instance selection
- data analysis
- domain adaptation
- positive examples
- sentiment classification
- probabilistic model
- data points
- knowledge discovery