Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment.
Di JinZhijing JinJoey Tianyi ZhouPeter SzolovitsPublished in: AAAI (2020)
Keyphrases
- text classification
- natural language
- information extraction
- data cleaning
- bag of words
- text mining
- text categorization
- language modeling
- language processing
- semantic features
- logic programming
- natural language interface
- high level
- semantic interpretation
- dialogue system
- document classification
- text documents
- n gram
- question answering
- training data