I guess OCR quality assessment score of each content item is stored somewhere. Can we provide this in the dataframe?