Indian SMS Dataset for OTP and Phishing Detection

A synthetic and labeled dataset of 73,470 Indian SMS messages. It contains labels for OTP detection, OTP intent classification into nine categories, and phishing detection. The dataset was created by gandharvbakshi and was last updated on December 5, 2025.

Use Cases

• Train OTP detection models based on the 'predicted_is_otp' label.
• Classify OTP intent into nine categories for message routing or analysis.
• Build phishing detection systems using the provided phishing labels.
• Research spam filtering techniques for SMS-based communication.

Strengths

• Contains 73,470 rows of labeled SMS data.
• Provides three distinct labeling tasks: OTP detection, OTP intent classification, and phishing detection.
• OTP intent is classified into nine specific categories.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is known, but the dataset's representativeness and potential geographic or source bias are not detailed.
• The synthetic nature of the data may not fully reflect real-world message distribution.

Provenance

Source

huggingface user gandharvbakshi

Collection Method

Synthetic generation and labeling.

Freshness

Last updated 2025-12-05 07:25:44.

Geography

India