A keyword-plus-cosine-similarity phishing detector reports 79.8 percent TF-IDF and 67.2 percent semantic accuracy, but uses the test set to pick its decision threshold.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Enhancing Phishing Detection in Financial Systems through NLP
A keyword-plus-cosine-similarity phishing detector reports 79.8 percent TF-IDF and 67.2 percent semantic accuracy, but uses the test set to pick its decision threshold.