Abdallah-Thuieb/social-spam-clean

social-spam-clean Arabic/English social media text dataset for spam detection (cleaned/standardized version of the merged dataset). This clean dataset focuses on keeping only the fields needed for modeling and analysis, with non-essential metadata removed where applicable.

Columns URL: original post URL (if available).

Content: unified text field used for training/inference (may be copied/filled from URL or Tweet_Text depending on your cleaning step).

Tweet_Text: tweet text (if available).

Spam_Ham: spam or ham label (if available).

Label meaning spam = spam/promotional content.

ham = normal content.

Notes This dataset is the cleaned output of the merged “social-spam” dataset: duplicate/irrelevant columns were dropped (e.g., Date/Time/Author if you chose to remove them), missing values were handled, and the text fields were normalized to make the dataset ready for training and evaluation.