SMS PHISHING DATASET FOR MACHINE LEARNING AND PATTERN RECOGNITION

The dataset is a set of labelled text messages that have been collected for SMS Phishing research. It has 5971 text messages labeled as Legitimate (Ham) or Spam or Smishing. It includes 489 spam messages, 638 smishing messages, and 4844 ham messages. This dataset contains raw message content that can be used as labelled data in Deep Learning or for extracting further attributes. The dataset contains extracted attributes from malicious messages that can be used for Classification of messages as malicious or legitimate. This dataset also includes python code that are used for extracting attributes. The data has been collected by converting the images obtained from the Internet to text using Python code. Attributes have been selected based on their relevance. The details of dataset attributes are given below:

LABEL- Classification label categorizing the message as ham, spam, or Smishing
TEXT- The raw content of the message.
URL- Gives out whether the message contains a URL or not.
EMAIL- Gives out whether the message contains an email id or not.
PHONE - Gives out whether the message contains a phone number or not.
Python code for extraction of the above listed dataset attributes is attached. The snapshot of this dataset is also attached. Frequency chart of the attributes are also attached.