Extracting features from the Kaggle/StumbleUpon evergreen classification dataset