Towards Automatically Creating Large Labeled Datasets for Training Question Domain Classifiers.
Leandro L. TavaresRenato Moraes SilvaTiago A. AlmeidaPublished in: IJCNN (2018)
Keyphrases
- training set
- training data
- labeled training data
- training dataset
- linear svm
- training and testing data
- training samples
- supervised learning
- training examples
- ground truth labels
- weakly labeled data
- manual labeling
- test set
- trained classifiers
- accurate classifiers
- decision trees
- domain specific
- training process
- linear classifiers
- support vector
- training set size
- classifier training
- sufficient training data
- class labels
- co training
- multi layer perceptron
- associative classifiers
- labeled images
- text classification tasks
- labeled data
- training and test data
- training stage
- naive bayes
- unlabeled training data
- class imbalanced
- test data
- classification accuracy
- boosted classifiers
- data sets
- unlabeled data
- discriminative classifiers
- unlabeled samples
- training instances
- text classifiers
- labeled examples
- svm classifier
- cross domain