How to use Classify Document and Extract Document Data Activities in Windows Template of Document Understanding

We already know that Classify Document and Extract Document Data Activities are being used in Cross-Platform template of Document Understanding.

What if we want to use them in Windows template?

Key problem - Classify Document and Extract Document Data Activity returns data in DocumentData Data type which is handled in Cross-platform template. Whereas windows platform uses ClassificationResult[] and ExtractionResult. In order to use DocumentData in windows template we would need to make some changes to the template.

Classify Document Activity

This activity will return output as DocumentData. To ensure this variable works with out argument (out_ClassificationResultsArray) , we need to convert DocumentData to ClassificationResults[].

With the help of above expression we can convert it.

Extract Document Data Activity

This activity will return output as IDocumentData<DictionaryData>. To ensure this variable works with out argument(out_ExtractionResults), we need to convert IDocumentData<DictionaryData> to ExtractionResult.

With the help of above expression we can convert it.

Advantage of using Classify Document and Extract Document Data Activities

1. This activities will consume ML skills from Document Understanding tab.

2. Like cross-platform template, It will eliminate the steps to create taxonomy and digitizing the document.

3. Eliminating Digitization would save some AI units.