We have an issue regarding the value of “Certainly Recognized Characters” when we process the documents in an “Unattended” workflow. We wish to process the documents with a 100% certainty of recognition of characters, apparently, we are only getting 91% in average.
We currently believe that when we set PDF text layer instead of OCR in the “Processing of PDF”, we will achieve the 100% value. But still we dont achieve the targeted 100% certainty.
Furthermore, does the priority of batch or the documents in the verification stage affects the certainty of recognition? As when the value of “Certainly Recognized Characters” is less than 100%, priority is shown as “NORMAL”. I already defined the priority as “Highest” in hte batch type properties, but why is it processed as “Normal”.
We are using a UIPATH-ABBYY Connector tool to process the documents. The version that we have is ABBYY Flexicapture 12_Distributed.
please noted that, you will never have 100% at least your PDF quality meet ABBYY requirement.
“Does the priority of batch or the documents in the verification stage affects the certainty of recognition?”
=>no. U set ur workflow as unattended, it’s mean will never go to verification stage.
I already defined the priority as “Highest” in hte batch type properties, but why is it processed as “Normal”.
=>where are you defiend it? project setting or batch setting?
“please noted that, you will never have 100% at least your PDF quality meet ABBYY requirement.”
==> is there a way to meet the ABBYY requirement for PDF quality? the documents we are processing are native pdf.
I defined the priority as “Highest” in both batch type setting and project setting.
I reviewed the manual for the ABBYY-UIPATH connector tool and ended up getting a 100% certainty of recognition of character when I selected the “For each image file” to create document in the Image Processing tab of Batch TYpe Properties. But with this setting, it seems we may have a problem with the output as ABBYY will process even if the Document Definition is not applicable.
I was using “Automatically when Document Definition is Applied”. Apparently, the certainty of recognition of characters becomes LESS than 100%. My client wants a 100% recognition so I switched to set “For each image file” to create document in the Image Processing tab of Batch Type Properties, then, the results were 100% recognized.
Based on @irahmat 's response, “U set ur workflow as unattended, it’s mean will never go to verification stage.”, I just found other way to show a 100% recognition of characters. By unchecking “Requires verification” of each field in the Document Section.