Document understanding questions

Hi guys,
It’s my first day of trying to use document understanding. All has been well until I landed here. Would anyone know why the items I selected on the first screenshot are not being displayed on the second screenshot for validation?
NB:This is for the second template. The first one worked perfectly.

Please help

Hi @J_Crosby ,

Could you select the Text Only View Option and check whether the values are properly present as put in the Keyword Classifier ?

This could also be the reason as it was trained or only has the data of one Template, the confidence gets balanced or increases when supplied or validated with Different sets of Templates at first and then check with newer templates.

Make sure you’ve clicked Configure Classifiers in the Classify Document Scope activity and checked the box for each document type you want it to classify for. If you don’t, it ignores what you’ve done in “manage learning.”

@postwick thanks,I have confirmed and they are all checked

@supermanPunch thanks, please direct me step by step where to find that. I have looked everywhere with no luck

@J_Crosby ,

Assuming Text Only View option was asked, you could check in the below documentation :

@supermanPunch its only one words that is being highlighted and marching with the ones put in the keyword classifier. I have tried with several other different documents and the result is the same. What could be the meaning of this?

@J_Crosby ,

As mentioned earlier, the supplied document types need to be also classified correctly and feed the corrected classification from Classification Station to the Train Extractors activity so that it is well trained with the different document formats and able to identify the different samples of different formats in the future as well.

The Below Video tutorial should help you understand on the Train Classifiers :

@supermanPunch thanks, am actually using Rakesh’s tutorial to learn.
Let me watch this one too. Thanks a lot for sharing

1 Like

What you’ve configured here are the keywords for it to identify that the document type is LPOs. It has done that correctly. If you also want it to extract those fields from the document, you need to go into Taxonomy Manager and add them as fields to the LPOs document type. Also, to see and edit the data it pulls for the fields you use a validation action (or present validation station if attended).

@postwick you nailed it :raised_hands:. This is exactly what I needed to know how to go about it. It worked. Thanks a lot.

1 Like

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.