The authors build the ICED benchmark of 440 cybercrime euphemism sentences and a context-augmented masked-language-model pipeline that detects held-out drug euphemisms with about 0.5 percent precision at top-20 retrieval.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Impromptu Cybercrime Euphemism Detection
The authors build the ICED benchmark of 440 cybercrime euphemism sentences and a context-augmented masked-language-model pipeline that detects held-out drug euphemisms with about 0.5 percent precision at top-20 retrieval.