Having the text in the database would allow the decision about whether to redact to be based on a dynamic whitelist which could change over time, without having to re-do the OCR. It would also mean a database would end up storing PII which is a consideration if it's ever to be accessed from outside the SDMZ.
howff
commented
1 year ago
Having the text in the database would allow the decision about whether to redact to be based on a dynamic whitelist which could change over time, without having to re-do the OCR. It would also mean a database would end up storing PII which is a consideration if it's ever to be accessed from outside the SDMZ.