Unsupervised Training Set Generation for Automatic Acquisition of Technical Terminology in Patents.
Alex JudeaHinrich SchützeSoeren BruegmannPublished in: COLING (2014)
Keyphrases
- training set
- supervised learning
- unsupervised learning
- training data
- data driven
- test set
- wide variety
- fully automatic
- semi supervised
- classification accuracy
- neural network
- domain specific
- active learning
- test data
- united states
- classification algorithm
- semi automatic
- cross validation
- terminology extraction
- training examples
- unlabeled data
- semi supervised learning
- information retrieval
- machine learning
- training samples
- nearest neighbor
- pairwise
- labor intensive
- intellectual property