You are viewing limited content. For full access, please sign in.

Question

Question

Quick Fields Capture Engine Processing

asked on April 15, 2014

The original problem has now caused other problems. I have limited experience with Quick Fields (8.3) and that's where this problem starts.

 

The basic process involved starts at a workstation where documents are scanned. The resulting documents are stored in a folder in the repository. There is another workstation used for nothing but running Quick Fields which picks up the documents and completes the process. The problem started when Quick Fields could not identify certain document classes. Pattern Matching is being used for identification of all four document classes. When I started, Invoices were the primary concern because when they came into Quick Fields they showed up in the document manager as unidentified documents. I got some help from Laserfiche Support and even sent a copy of the configuration. The reply I got back was that the configuration could have never worked and the Identification process needed to be done with Zone OCR. I worked on trying to convert those processes without any success. I went back to working on the Pattern Matching solution and had some success. Then we ran a batch to test and Invoices went through but Packing Lists were not being identified. I tweaked the pattern that was being used in the token dialog and what I did passed the test. It did not work during the process however.

 

My immediate problem is with documents sitting in the Document Manager that I have to do something with. I have unidentified Packing Lists that I can right-click and assign to a document class but when I do that they go somewhere and I'm not sure where that is. I also have documents above the unidentified section that have icons that look like a document with a red circle in the bottom right corner. I've been told to get those unidentified Packing Lists into a specific folder in the repository with the matching Invoice and Sale Order. I'm not sure at this point how to get that done.

0 0

Answer

SELECTED ANSWER
replied on July 23, 2014

If a response answered your question, please click "This answered my question" to let us know.

0 0

Replies

replied on April 15, 2014

Maybe a screenshot will help us out.  I am not sure what the red circle icons you are referring to.  When you assign a document to a class it will move itself out of Unidentified and into the Document Revision area.  Depending on your class settings it could go to a folder with lookup, zonal ocr, or pattern matching tokens.  Point is, when you manually assign it a class it moves up to the Document Revision.  It does not delete it or store it so it should be there.  You can manually change the path once you find it so you can store the docs and then get them.

1 0
replied on April 15, 2014

I have attached a screenshot of the Document Revision pane. I cut off the top to hide the customer name.

DocRevision.jpg
0 0
replied on April 15, 2014

Those red icons generally mean that there is a required field that is not filled out. If the document was not identified originally and you classified it, there will be no data to fill the fields. Also, when you classify a document, it goes from the unidentified node to slotting into one of the folders up above.

 

The overall issue is probably the quality of your session configuration and processes, unless your documents are extremely difficult to identify.

0 0
replied on April 15, 2014

From your screenshot I am guessing that the ones you assign to the class manually are located right above unidentifed in Sales Order\SO% (expand the folder).  When you assign a document to a class manually and non of the page processing are reading it is going to assign it the tokens instead which is why the folder is called SO% and not SO19 like the others.  

 

The red exclamation points are just errors on the document.  Select them and then over on the rights see what's up.  It could be a missing required field or some other metadata problem (not the correct format, etc).

0 0
replied on April 15, 2014

You are correct about the unidentified documents moving after the class is assigned. I can now see what data is missing for each of the documents. Do I try to manually fix these and put in the missing data or is there a way to fix this?

0 0
replied on April 16, 2014

Once a document is moved into the Identified documents area you are going to have to manually fix the problems.  Fill in the path, change the name, fix the metadata, etc.  Another way is to delete them and scan them again.  If you do this you may get the same result so you would need to change/fix the zones.  This is one of the many reasons us as a VAR and our clients do a barcode and lookup process instead of zonal OCR.  The accuracy and consistency is just not there (in our situations).

0 0
You are not allowed to follow up in this post.

Sign in to reply to this post.