Financial scams detection dataset

Published: 22 October 2025| Version 2 | DOI: 10.17632/znsk27yk3h.2
Contributors:
,

Description

This dataset contains real financial scam and non-scam (ham) messages in both English and Bangla, collected from university students and general people through Google Forms, handwritten submissions, and direct observation of mobile SMS, chat messages, and emails with proper consent. The data includes various types of financial scam contents such as fake bKash/Nagad offers, lottery winnings, loan and job scams, OTP frauds, banking and account-related phishing texts, as well as genuine everyday messages for comparison. All personal or sensitive information like names, phone numbers, account numbers, or email addresses has been removed or anonymized to ensure privacy and ethical use. Each message is manually labeled as either “scam” or “ham,” and the dataset is structured in CSV format for research purposes related to financial fraud awareness, analysis of scam message patterns, and multilingual (Bangla-English) text study.

Files

Institutions

  • International University of Business Agriculture and Technology

Categories

Data Science, Natural Language Processing, Machine Learning, Bankruptcy Fraud

Licence