Tamp-X: Attacking explainable natural language classifiers through tampered activations.
Hassan AliMuhammad Suleman KhanAla I. Al-FuqahaJunaid QadirPublished in: Comput. Secur. (2022)
Keyphrases
- natural language
- decision trees
- training data
- machine learning algorithms
- support vector
- training set
- semantic analysis
- knowledge representation
- linear classifiers
- feature selection
- classification systems
- feature set
- natural language processing
- multiple classifiers
- naive bayes
- natural language generation
- ensemble learning
- semantic interpretation
- classifier ensemble
- machine learning
- natural language interface
- roc curve
- classification method
- classification algorithm
- test set
- classification models
- dialogue system
- svm classifier
- question answering
- classifier combination
- image classification
- information extraction