A two-phase hybrid of semi-supervised and active learning approach for sequence labeling.
Hamed HassanzadehMohammadreza KeyvanpourPublished in: Intell. Data Anal. (2013)
Keyphrases
- semi supervised
- sequence labeling
- active learning
- named entity recognition
- labeled and unlabeled data
- semi supervised learning
- conditional random fields
- labeled data
- unlabeled data
- dependency parsing
- supervised learning
- unsupervised learning
- generative model
- multi view
- discriminative models
- pairwise
- random sampling
- co training
- structured prediction
- hedge detection
- crf model
- training examples
- gaussian process
- machine learning
- information extraction
- pairwise constraints
- learning algorithm
- random fields
- generalization error
- probabilistic model
- learning process