Domain Classification-based Source-specific Term Penalization for Domain Adaptation in Hate-speech Detection.
Tulika BoseNikolaos AletrasIrina IllinaDominique FohrPublished in: CoRR (2022)
Keyphrases
- domain adaptation
- target domain
- multiple sources
- cross domain
- document classification
- source domain
- training samples
- domain specific
- transfer learning
- labeled data
- sentiment classification
- classification accuracy
- co training
- semi supervised
- text classification
- semi supervised learning
- feature vectors
- machine learning
- feature extraction
- classification algorithm
- supervised learning
- source and target domains
- decision trees
- feature selection
- support vector machine
- training set
- support vector
- training data
- class labels
- test set
- knowledge transfer
- error rate
- naive bayes
- text categorization