Go Simple and Pre-Train on Domain-Specific Corpora: On the Role of Training Data for Text Classification
3886Knowledge Triples
| Source Entity | Relationship | Target Entity |
|---|---|---|
|
BERT
Method |
performs |
Text Classification
Task |
|
Domain-Specific Unlabeled Data
Material |
enhances |
BERT
Method |
|
Domain-Specific Unlabeled Data
Material |
enhances |
Word Embeddings
Material |
|
FastText
Method |
performs |
Text Classification
Task |
|
Pre-trained Language Models
Method |
supports |
Text Classification
Task |