Smishing AZ SC: Azerbaijani SMS Messages Classified as Ham, Spam, or Smishing

An SMS dataset for multi-class classification tasks in the Azerbaijani language, distinguishing between legitimate messages (ham), spam, and smishing (SMS phishing). The dataset was created by Vusal Shahbazov and published in the journal Problems of Information Technology in 2026. It was last updated on the Hugging Face platform on May 8, 2026.

Use Cases

• Train a multi-class text classifier based on the described categories of ham, spam, and smishing.
• Benchmark natural language processing models for Azerbaijani language understanding based on SMS text data.
• Study the linguistic patterns of smishing (SMS phishing) attacks in a specific language context based on the dataset's focus.
• Develop filters or alert systems for mobile messaging platforms based on the classification of Azerbaijani SMS messages.

Strengths

• Dataset is explicitly designed for a multi-class classification task with three distinct categories (ham, spam, smishing).
• The dataset has a formal academic citation, indicating peer-reviewed publication in a journal (Problems of Information Technology, Volume 17, Number 1, 2026).
• It focuses on the Azerbaijani language, which may address a gap in resources for this linguistic domain.

Limitations

• The number of rows, columns, and overall data size are unknown, making it difficult to assess scale and suitability.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality and balance between classes require manual inspection after download.

Provenance

Source

Vusal Shahbazov via Hugging Face.

Collection Method

Likely collected and annotated SMS messages, but the specific gathering method is not detailed in the provided input.

Freshness

Last updated 2026-05-08 12:32:52; freshness should be verified.

Geography

Azerbaijan (inferred from title and description focus on Azerbaijani language).