The dataset contains text with binary labels for suicide or non-suicide.
The dataset was cleaned and following steps were applied
Converted to lowercase
Removed numbers and special characters.
Removed URLs, Emojis and accented characters.
Removed any word contractions.
Remove any extra white spaces and any extra spaces after a single space.
Removed any consecutive characters repeated more than 3 times.
Tokenised the… See the full description on the dataset page:
https://huggingface.co/datasets/vibhorag101/suicide_prediction_dataset_phr.
