one document type vs. multiple document types

[siddharthjindal]

Hi,

I'm trying to train Decipher to read information from remit slips (tabular data). However, the remit slips may differ in their format slightly basis the bank. I want exactly the same column data from each remit slip irrespective of format (header name, column positioning vary).
So should i create multiple document types like remit slip-bank A, remit slip--bank B; or should i create just one document type - calling it remit Slip and then classify all remit slips as one document type?

Regards
Sid

[BenLyons · Accepted solution]

Hi Sid,

If you're looking to extract the same values, then you should only need 1 DFD and Document Type. This is the main benefit of using Decipher that only need to configure it once and the training will then manage itself by layout.

Thanks

Ben

[lookman]

Hi Sid.

To add onto Ben's answer, you can have multiple sample headers for a particular field so if doc A has a different header than doc B, include them both for the field.

jack

[siddharthjindal]

Thank you Ben and Jack. 🙂