Text classification using label names only: A language model self-training approach

Keyphrases

• Training Approaches 100%
• Language Model 100%
• Text Classification 100%
• Self-training 100%
• Small Sets 50%
• Training Data 50%
• Benchmark Dataset 50%
• Real Application 50%
• Linguistic Knowledge 50%
• Classification Model 50%
• Learning Model 50%
• Classification Methods 50%
• Unlabeled Data 50%
• Semantically Related Words 50%
• Sentiment Classification 50%
• General Linguistics 50%
• Document Classification 50%
• Train Classification 50%
• Representation Learning 50%
• Neural Language Models 50%
• Topic Classification 50%
• Learning from Unlabeled Data 50%

Computer Science

• Language Modeling 100%
• Text Classification 100%
• Labeled Document 100%
• Unlabeled Data 66%
• Representation Learning 33%
• Training Data 33%
• Classification Method 33%
• Classification Models 33%
• Real Application 33%
• Labeled Example 33%
• Sentiment Classification 33%
• Document Classification 33%