Loading...
Loading...
Contributors:
Dataset URI:
10.5281/zenodo.23338781The three external corpora of the PhishVN URL-benchmark study, in the exact rows the study used: class-balanced 1:1 subsamples of PhiUSIIL (20,000 rows, UCI, label polarity converted so that 1 is phishing), ISCX-URL2016 (15,894 rows) and PhishStorm (20,000 rows), the last two in the phishing/legitimate form redistributed with the GramBeddings repository. Each row carries the host the study featurises, 25 lexical features in the CompPhish schema, computed with the public-suffix list of the library version used, and the label. The subsamples were not drawn by a seed, so these files define the transfer experiments. Together with the PhishVN corpus (Mendeley Data, DOI 10.17632/b97hxbxtpd.4), they let the notebook notebooks/url_benchmark/url_benchmark_walkthrough.ipynb in https://github.com/vuthainguyen1602/phishvn recompute every number of the paper. Cite the original datasets when using the rows (see README.md). Files (see README.md for the column list and the citation terms): phiusiil_compphish.csv: 20,000 (10,000 phishing, 10,000 legitimate) rows, PhiUSIIL Phishing URL Dataset, UCI Machine Learning Repository (CC BY 4.0), . The original codes 1 as legitimate; here `label` 1 is phishing.iscx_compphish.csv: 15,894 (7,947 each) rows, ISCX-URL2016, Canadian Institute for Cybersecurity, , phishing and benign portions, in the phishing/legitimate form redistributed with the GramBeddings repositoryphishstorm_compphish.csv: 20,000 (10,000 each) rows, PhishStorm, Aalto University, , in the form redistributed with the GramBeddings repository SHA256SUMS lists the SHA-256 of every file.
Keyword(s):
benchmark; cross-dataset transfer; ISCX-URL2016; phishing detection; PhishStorm.
MeSH Term(s):
Grants and Funding:
No Grants or Funding Data Available
Associated Publication(s):
No Associated Publication Data Available
Supplementary Data:
Related Dataset(s):
No Related Dataset Data Available
The National Library of Medicine (NLM) provides information about biomedical datasets. Inclusion on an NLM Web page does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: NLM Web Policies (nih.gov)