Intelligent OCR Activity


Can somebody please explain briefly, what does Classify document activity do.

I went through the documentation provided by UiPath, but I didn’t get much information.

Can someone help me?

Thanks in Advance!!

Hey @kantheshm

Well in Abbey Flexicapture let you classify documents and detect their types and variants.
ABBYY FlexiCapture is able to identify the type of a document by finding and applying a matching Document Definition(A Document Definition describes the location of document elements and indicates fields to be used in data extraction. Document Definitions can only be created in Administrator mode. Document Definition can contained either a FlexiLayout or a fixed form template).

Classification can detect document types based on various image elements, such as image patterns, separators, barcodes, large titles, and full-text content. Once the program has detected the document type (or identified several possible types) using classification, it tries to apply only those FlexiLayouts that correspond to the detected types.

In short Classification in abbey to identify that input document belong to an Invoice Document type category, Letter or else.


Hi @aksh1yadav,

Thank you for your Replay. It helped me a lot.

And I am also struck with another issue can you help me…

Can we create one FCDot file for a set of Different Document with specified field??
And also I want that same FCDot file to identify another set of the document with different Pattern…

Is it possible to do that in FlexiCapture?
If so, can please suggest me??



Yes, I have created one .fcdot file for some 20 different documents and its working flawlessly for me.

Create different layouts(.afl files) for different documents, into one document definition (.fcdot). and Use that document definition into the Intelligent OCR Activities.


Hey @nshrimali,

Thank you for your replay.

Okay i can do that.But if we get another set of new document then once again we have to build it from scratch right??


Hey @nshrimali,

How did you create one FCDot file for 20 different documents??

i have created different layouts(.afl) for different document.But after that i am trying to create one fcdot in document definition but i couldn,t do that…

Can please help me how did you do that??


Hi @nshrimali
did u use paid version of abbey flexicapture?
or the community one
will this abbey be helpful in scrapping unstructured data from emails also
if u know smthng about abbey.please let me know




I have used trail of Abbyy, since we are partner with UiPath i got a trial version of ABBYY.

It works fine for unstructured data, but as usual, you know how OCR works when image resolution varies.
Same difficulty is faced in Abbyy also.

But compared to other OCR ABBYY is the best.

Try it out once!!

Happy Learning!!

i don’t have enterprise edition
is abbey available for trial version also?
pls let me know if it is?
i hv unstructured data in email body.i hope its useful for dat scenario as well.

thanks buddy

Hi @niteshbutola5,

Sorry for the delay!!

If you want to see the proper automated flow go with an enterprise edition.Anyways trail of 60days is available.

Right now i dont have any idea about trail version of ABBYY you have to check with UiPath team for that.


1 Like


Assuming emails are not images and are in unstructured format, if you already have an email body, then I am not sure how Abbyy makes sense in extracting the data out of that.

ABBYY is useful when you are extracting the information out of scanned PDF or image, like I used an licensed version for invoice processing(Scanned PDF). For your use case, I would recommend you trying NLP(Natural Language Processing). You can try LUIS and trails of these are also available for a limited period of time and with limited computation.

1 Like

Hey @nshrimali

thanks for ur valuable suggestions.
can u pls share some resources of NLP,LUIS that u suggested to use if u hv…/
M beginner.i don’t hv any idea how to integrate NLP with uipath.
if u know Please share with me as well

thanks again buddy

Hey Nitesh,

Check out the below link, this is all you need to do to train on LUIS for extracting the relevant information.

Also it can be easily integrated with UiPath using the API calls.

Happy Learning!!!..