02-01-24 10:38 PM
Hi,
I am struggling with one more thing with DFD in parallel and due to urgency of the problem I am not able to create new ticket and wait for response.
I have to extract 2 things - Invoice number and title of the Invoice (word "Invoice")
I created 2 fields:
InvNo with sample headers like 'Invoice, Invoice number' etc.
and
InvTitle as a selectable field with option 'Invoice' and 'Other' as default.
My problem is that the InvNo field "takes" the "Invoice" title as its header to find the invoice number and then the "Invoice" value is not assigned to the "InvTitle" field.
How to make it works?
04-01-24 10:09 PM
Hi Wojciech,
I had faced the similar issue in my current process. Here is how I resolved the issue.
04-01-24 10:26 PM
Hi Athiban,
thanks for your input. In my case I cannot define specific version trigger as vendor because I have too many invoice vendors to proceed and every month new templates appears.
I am wondering if I can create Specific Version of DFD based on rule 'when some field value is '' (empty)' but I dont think it would work because problem is that I have to use same word on invoice for two differenct purposes - value of field A and sample header of field B.
I was thinking to create 2 batch types, with 2 different DFDs, one for extracting this word "INVOICE" as a value of one field and another DFD and batch type for using word "INVOICE" as a sample header for invoice number. But I think that it is far away from good practices and it multiplies time complexity.
08-01-24 08:12 AM
Hi Wojciech,
I understand your point, though I would still advise against using a keyword that appears multiple times in a document. It's best to allow Decipher to use it's training data ML to learn and capture these fields. Hopefully this can be demonstrated in your session today.
Using 2 DFDs for the same document doesn't tend to work as they will use the same training template, these are not currently locked to a specific DFD. Training data is a 'pool' of document training templates shared by all DFDs.
Any low confidence document or document that fails validation will be held for manual verification, even after auto-skip is enabled. To ensure all low confidence fields 'trigger' this behaviour, add the "Required" flag. Any low confidence/blank fields with this flag will not be automatically processed.
Thanks
01-02-24 05:22 PM
Hi Wojciech,
Your urgent issue is about a month old but thought I would reply anyway as a matter of general interest. If your field invNo is only populated when there is an invoice number then the field invTitle could be set to "Invoice" (instead of "Other") using a formula instead of trying to read the word "Invoice" from the document.
In the invTitle field add a formula to determine if invNo is blank:
IF(EQ(invNo, ""), "Other", "Invoice")
Don't forget to add "AutoCalculate" to the flags of invTitle.
------------------------------
Stuart Mar
------------------------------