Active Learning Yields Better Training Data for Scientific Named Entity Recognition.
Roselyne TchouaAswathy AjithZhi HongLogan T. WardKyle ChardDebra AudusShrayesh PatelJuan de PabloIan T. FosterPublished in: eScience (2019)
Keyphrases
- named entity recognition
- active learning
- semi supervised
- training data
- labeled data
- supervised learning
- training examples
- unlabeled data
- training set
- information extraction
- learning algorithm
- named entities
- natural language processing
- generalization error
- semi supervised learning
- maximum entropy
- conditional random fields
- text summarization
- domain adaptation
- data sets
- relation extraction
- machine learning
- sequence labeling
- test data
- class labels
- classifier ensemble
- annotated corpus
- unsupervised learning
- generative model
- pairwise
- decision trees
- classification accuracy
- artificial intelligence
- training samples
- prior knowledge
- question answering
- learning process
- data analysis
- data mining