R-AT: Regularized Adversarial Training for Natural Language Understanding.
Shiwen NiJiawen LiHung-Yu KaoPublished in: EMNLP (Findings) (2022)
Keyphrases
- natural language understanding
- text understanding
- semantic analysis
- knowledge representation
- natural language processing
- natural language
- language understanding
- semantic representations
- training set
- joint inference
- abductive reasoning
- dialogue system
- least squares
- multi agent
- spoken dialog systems
- chinese word segmentation
- hidden markov models
- intelligent agents
- ground truth
- relational databases
- expert systems
- databases