Targeting the Most Important Words Across the Entire Corpus in NLP Adversarial Attacks.
Reza MarzbanJohnson ThomasChristopher CrickPublished in: ICAISC (2) (2021)
Keyphrases
- english words
- word frequencies
- natural language processing
- word sense disambiguation
- text corpora
- multiword
- training corpus
- unknown words
- text corpus
- word pairs
- linguistic information
- part of speech
- linguistic knowledge
- text mining
- hand crafted
- word sense
- natural language
- broad coverage
- coreference resolution
- pos tagging
- reference resolution
- n gram
- countermeasures
- document level
- person names
- text analysis
- information extraction
- linguistic features
- natural language text
- computational linguistics
- keywords
- question answering
- pos taggers
- lexical semantics
- text documents
- named entities
- conversational speech
- lexical features
- word co occurrence
- semantic roles
- linguistic analysis
- free text
- noun phrases
- wordnet
- ontology learning
- semantic relations
- statistical machine translation
- spontaneous speech
- machine translation
- dependency parsing
- word segmentation
- grammar induction
- word frequency
- lexical resources
- language model
- world knowledge
- textual entailment