@SrenivasanKanna ,Yes, You are correct. The Action Center Trained document values are nothing but the Validated extraction results that we receive. As you can see from the Screenshot in the Trainer Extractor Scope, Human Validated data is the data output after the Action Center task. So we use the Machine Learning Extractor Trainer Activity to train our custom model with the validated data from the users.
Thank you so much SuperMan !!! one more question on Action Center. While moving cases into Action Center? Can we validate all the pending documents later? which means can the bot pick the next document or is it waiting for User validation to pick next case?
@SrenivasanKanna , This part actually depends on the way the Solution is Designed and what is the actual scenario.
Generally, If a Parallel For Each is used, it will Extract and send the Documents to Action Center at once, so the User/Admin can assign it to himself/others so that the document validation can be done parallelly, then after user submits the data, if the remaining operations are background then it will be completed for the submitted value/case.
Here, if the remaining operations contain UI Operations then it is better to keep it out of Parallel For Each, which would mean it would wait for all the document validations from Action center and then continue the UI Operations for each of them.
There are other Strategies when there are Multiple Bots being used, You could use a Queue to add the validated data to the Queue, so another Bot picks it up and performs the remaining operation.