StreminiAI: 39,000 Labeled URLs for Phishing and Malware Detection

39,000 labeled URLs with extracted features for cybersecurity classification. The dataset is hosted on Kaggle and appears designed for machine learning tasks in threat detection. Its creation date, author, and specific feature details are not provided in the available metadata.

Use Cases

• Training binary classifiers to detect phishing URLs based on extracted features.
• Training multi-class models to distinguish between benign, phishing, and malware-hosting URLs.
• Feature engineering analysis to identify which URL characteristics are most predictive of threats.
• Benchmarking new detection algorithms against a labeled dataset of URLs.

Strengths

• 39,000 labeled instances provide a substantial base for model training.
• Focus on a specific, high-impact cybersecurity application: URL threat detection.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is known, but data freshness, source, and collection methodology are unknown.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Kaggle

Collection Method

Likely gathered through web crawling and labeling for security research.

Time Range

null

Freshness

Last update date is unknown; freshness unverified.

Geography

null