Domain Classification-based Source-specific Term Penalization for Domain Adaptation in Hate-speech Detection.
Tulika BoseNikolaos AletrasIrina IllinaDominique FohrPublished in: COLING (2022)
Keyphrases
- domain adaptation
- target domain
- multiple sources
- cross domain
- source domain
- document classification
- domain specific
- transfer learning
- text classification
- training samples
- labeled data
- machine learning
- sentiment classification
- co training
- decision trees
- semi supervised
- classification accuracy
- feature space
- support vector machine
- semi supervised learning
- training data
- feature representation
- recommender systems
- feature vectors
- classification algorithm
- unlabeled data
- data sets
- multi task
- test data
- feature set