Multi-domain Adaptation for Statistical Machine Translation Based on Feature Augmentation.
Kenji ImamuraEiichiro SumitaPublished in: AMTA (1) (2016)
Keyphrases
- domain adaptation
- cross domain
- labeled data
- semi supervised
- manifold alignment
- transfer learning
- multiple sources
- target domain
- test data
- covariate shift
- sentiment classification
- feature representation
- feature vectors
- unlabeled data
- document classification
- feature set
- data sets
- text classification
- knowledge transfer
- dimensionality reduction
- supervised learning
- databases