Xtrata always classifies documents

Xtrata always classifies documents

Postby hanneseklund » Tue Mar 20, 2007 3:49 am

Hi,

I have a problem with Xtrata, it always classifies the document, even if the minimum confidence is not reached!

I have configured 10 form types, and as long as I scan one of theese forms, everything works fine, the document is classified correctly (>90% confidence).

But if I scan an unknown document, and get a low confidence rating, the document is still classified.
The Xtrata log says:
"Document 1 was classified with confidence 44.52%"

I have "Min. confidence" set to 70%, but it seems to be ignored.

I am using Xtrata 1.7.208.0

Anyone experienced the same thing ?

/Hannes
hanneseklund
Participant
 
Posts: 6
Joined: Fri Mar 09, 2007 5:58 am

Postby wolfgang.radl@kofax.com » Tue Mar 20, 2007 4:06 am

Do you have a standard formular type defined (within the Xtrata Setup)?
Kind Regards,
Wolfgang Radl
KOFAX Sales Engineer
wolfgang.radl@kofax.com
Participant
 
Posts: 453
Joined: Sun Oct 09, 2005 10:52 pm
Location: Vienna, Austria

Postby hanneseklund » Tue Mar 20, 2007 4:49 am

I have tried with and without a "default form type", the result is the same.

(In my final setup I do want to have a default form type)

/Hannes
hanneseklund
Participant
 
Posts: 6
Joined: Fri Mar 09, 2007 5:58 am

We had such problem

Postby pleshkov@prae.ru » Wed Mar 28, 2007 8:43 am

In our case in Russia we have delimiter - comma instead of dot. So we changed regional settings to english and everything start to work fine.
pleshkov@prae.ru
Participant
 
Posts: 14
Joined: Sat Oct 29, 2005 12:09 pm

Postby petkove » Sun Apr 15, 2007 11:23 pm

And what is the workaround?
I don't think it is acceptable for customers to change the regional settings...
Is there any config file to change this?
tnx for the previous post btw
petkove
Participant
 
Posts: 2
Joined: Sun Apr 15, 2007 11:20 pm


Return to Ascent Xtrata General Discussion

Who is online

Users browsing this forum: No registered users and 1 guest