- I have two types of document , lets say A and B.
- In a path these documents will be kept.
- All the documents path will be added to queue.
- In Document Understanding , how the bot will distinguish classify and extract ?
The is a documnet classification activity and different types of classifiers are present which will clasify based on what and how you configure
And coming to extraction again is based on the classification different details can be extracted
https://docs.uipath.com/document-understanding/docs/data-extraction-overview
Cheers
Hi @Ritaman_Baral ,
From your initial Statements, we understand that you have already assessed the requirement and found out that there are two types of Document.
So The Taxonomy would be created accordingly which will have two types to Distinguish.
You can then follow the Methods mentioned above and Train the Classifiers using the The Documents available using either of the below activities in Classify Document Scope
activity :
- Keyword Based Classifier
- Intelligent Keyword Classifier
There is also a way to use Machine Learning Classifier if you’re Document Types fall under it’s already trained document types, else a re-training of the Classifier would be required.
Next Steps, for the Extraction, the Classification results are given to the Data Extraction Scope
where you would have the Extractor activities configured : Form Extractor
, Regex Based Extractor
and Machine Learning Extractor
Over here too, the extractors to be used are dependent on the document types.
There are already Out-Of-the-Box ML Packages that you could use, or you would have to Train the Document Understanding Model by Performing Data Labelling for your custom Document types and Deploy the Model.
Also take a look at the Document Understand Process Template which is also available within the Studio Templates, it should answer most of Straight forward implementations.
Let us know if you are still looking for more specific answers.
This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.